Qba 221-Quantitative Business Analysis

MODULE IN QBA 221 (QUANTITATIVE BUSINESS ANALYSIS) – BSAB
2A & B
THE INTRODUCTION:
THE OVERVIEW:
INSTRUCTION TO THE USER:
THE OBJECTIVES:
LEARNING ACTIVITIES:
LESSON 1: SYSTEM OF LINEAR EQUATIONS AND INEQUALITIES
PRE-TEST:
Direction: Answer the question comprehensively and concisely.
1. Define the systems of linear equation and inequality.
2. Explain the types of systems of linear equations.
3. Explain the different methods of solving systems of linear equations.
4. Define inequality.
5. Describe the different properties of inequality.
6. What are the types of Inequalities?
Learning Activities:
The learner shall be able to:
1. Solve linear equations in one or two variables;
2. Graph the solution set of linear equations;
3. Compute the solution set of linear inequalities in one or two variables;
4. Graph the solution set of linear inequalities; and
5. Solve systems of linear equations and linear inequalities.
TYPES OF SYSTEMS OF LINEAR EQUATIONS:

 Linear equation – is an algebraic equation in which each term is either a constant or
a product of a constant and (the first power of) a single variable, mathematically it is
represented by ax + by = c where a, b, and c are constant while x and y are the
unknown variables.
 System of Linear – is the common point (ordered pair) of any two given linear
equation
a1 x b1y = c1 ----------------- First equation
a2 x b2y = c2 ----------------- Second equation
 Standard equation – an equation of the form, ax + by = c, where a, b ≠ 0 of the first
degree in x and y.
A. Inconsistent. An inconsistency has a solution of empty sets and the graphs consist of
a1 b 1 c1
two parallel lines. It can be determine using the ratio test = ≠ .
a2 b 2 c2
Zero solutions
Example: y=−2x+4y=−2x−3y=−2x+4y=−2x−3
B. Dependent. A dependent solution contains an infinite number of solutions and the

a1
graphs are two coincident lines and can be determine using the ratio test =
a2
b1 c 1
= .
b2 c 2
Infinitely many solutions:
Example: y=−2x−4y+4=−2xy=−2x−4y+4=−2x
C. Consistent. A consistent contains only one solution, and

the graphs are two lines, which intersect at one point. It can be easily determine
a1 b1
using the ratio test ≠ .
a2 b2
One solution:
Example: y=0.5x+2y=−2x−3y=0.5x+2y=−2x−3
 Methods of Solving System of Linear Equations:
After identifying that the system of linear equation is consistent, we can compute for
solution of the inserting lines.
The different methods of solving the solution of linear equation involving two unknown
variables:
1. Elimination Methods by Addition or Subtraction
Steps in Solving System of Linear Equations Elimination Method:
a. If the coefficients of the variables we choose to eliminate contain the same
coefficient, we proceed to step 2. In cases that the coefficients are not the same, we
multiply one equation or both by a number, which will make the coefficients of the
variables equal.
b. Add or subtract the resulting equations and solve for the value of the remaining
variables.
c. Substitute the value obtained in step 2 to get the value of the second variable to any
of the original equation.
d. Check by substituting the values of the variables in the original equations.
Example 1
4a + 5b = 12,
3a – 5b = 9
Solution
Since the coefficients b are the same in the two equations, we vertically add the terms.
4a+3a) +(5b – 5b) = 12 + 9
7a = 21
a = 21/ 7
a=3
substitute the obtained value of a=3 in the equation the first equation
4(3) + 5b = 12,
12 + 5b = 12
5b = 12-12
5b =0
b = 0/5 = 0
Therefore, the solution is a =3 and b = 0.
Example 2
Solve using elimination method.
2x + 3y = 9 ———–(i)
x – y = 3 ———–(ii)
Solution
Multiply the two equations by 2 and perform subtraction.
2x + 3y = 9
(-)
2x – 2y = 6
-5y = -3
y = ⅗ = 0.6
Now substitute the obtained value of y in the second equation: x – y = 3
x – 0.6 = 3
x = 3.6
Therefore, the solution is: x = 3.6 and y= 0.6
2. Substitution Method
Steps in Solving System of Linear Equations by Substitution Method
a. Choose the simpler equation and solve for one variable in terms of the other.
b. Using the other equation, substitute the expression obtained in step 1 to find the
value of the other variable.
c. Substitute the value obtained in step 2 to get the value of the second variable to any
of the original equations.
d. Check by substituting the values of both variables in the original equations.
Example 1.
The solution is (1, -5)

Example 2.
3. E
quivalent Equations:
Steps in Solving System of Linear Equation by Equivalent Equations
a. Choose one variable and solve for one variable in terms of the other.
b. Form a new equation from the equivalent equations and solve for the unknown.
c. Substitute the value of the first variable in step 2 to get the value of the second
variable.
d. Check by substituting the values of both variables obtained in the original equations.
Example 1.
Given the equations:
-3x + 12y = 15 ------------------- equation 1
7x - 10y = -2 --------------------- equation 2
Start with the first set. To solve two equations with two variables, isolate one variable and
plug its solution into the other equation. To isolate the "y" variable:
-3x + 12y = 15
-3x = 15 - 12y
x = -(15 - 12y)/3 = -5 + 4y (plug in for "x" in the second equation)
7x - 10y = -2
7(-5 + 4y) - 10y = -2
-35 + 28y - 10y = -2
18y = 33
y = 33/18 = 11/6
Now, plug "y" back into either equation to solve for "x":
7x - 10y = -2
7x = -2 + 10(11/6)
Working through this, you'll eventually get x = 7/3.
Example 2.
Equivalent equations with two unknown variables:

We have two unknown variables in both equations. You can only solve it simultaneously if
the roots of both equations are the same. This means that the line of both equations
intersects at a specific point.
Considering equation i and writing it in terms of x:
Substituting the value of x
Since, therefore:
Hence, we found that the value of y is - but still we have one variable remaining and that is
x. Now, we will substitute the value of y:
Since, therefore:
In conclusion, at the value of x for the both equivalent equations are . Let's
verify by substituting the value of x and y in both equations:

When:

When:
4. Cramer’s Rule
Steps in Solving System of Linear Equation by Cramer’s Rule
a. Determine the first and second equation.
b. Collect the values of a, b, and c of the two equations.
c. Substitute the values obtained in the formula.
d. Check by substituting the values of the variables in the original equations.
Evaluating the Determinant of a 2×2 Matrix
Example 1.
Find the determinant of the given 2 x 2 matrix.
Using Cramer’s Rule to Solve a System of Two Equations in Two Variables

To understand Cramer’s Rule, let’s look closely at how we solve systems of linear
equations using basic row operations. Consider a system of two equations in two variables.
We eliminate one variable using row operations and solve for the other. Say that we
wish to solve for x. If Equation 9.8.2 is multiplied by the opposite of the coefficient of y in
Equation 9.8.1, Equation 9.8.1 is multiplied by the coefficient of y in Equation 9.8.2, and we
add the two equations, the variable y will be eliminated.
Solve for x:
Solve for y &
eliminate x:
Solving for y gives
Notice that the denominator for both x and y is the determinant of the coefficient matrix.
We can use these formulas to solve for x and y, but Cramer’s Rule also introduces new
notation.
 D: determinant of the coefficient matrix
 Dx: determinant of the numerator in the solution of x
 Dy: determinant of the numerator in the solution of y
The key to Cramer’s Rule is replacing the variable column of interest with the constant
column and calculating the determinants. We can then express x and y as a quotient of two
determinants.
Cramer’s Rule for 2 x 2 Systems
Cramer’s Rule is a method that uses determinants to solve systems of equations that have
the same number of equations as variables.
Consider a system of two linear equations in two variables.
The solution using Cramer’s Rule is given as:

Example 2:
Solve the following 2 x 2 system using Cramer’s Rule.
Solution:
Solve for x:
Solve for y:
The solution is (2, -3).

Example 3:
Find the determinant of the 3 x 3 matrix
Example 3:
Find the solution to the given 3 x 3 system using Cramer’s Rule.
Solution:
The solution is (1, 3, -2).
5. Graphical Method
Steps in Solving System of Linear Equation by Graphical Method
a. Solve the two-given equation for x and y values (coordinates).
b. Graph each equation on the same Cartesian Coordinate Plane.
c. Find the point of intersection. The point of intersection is the solution set.
d. Check by substituting the obtained values of the variables in the original equations.
Example 1:
Solve the following system of equations graphically: y=-x+4 and y=x-2
STEP 1: The given equations represent the equations of lines. We have to find the
intercepts of the graph of the given equations and then use them to graph the equations.
First, find the x-intercept of y = –x + 4 by substituting 0 for x.
Now, replace y with 0 to determine the y-intercept.
The x-intercept is (0, 4) and the y-intercept is (4, 0).

STEP 2: Plot the intercepts and join
them using a straight line to get
the graph of y = – x + 4.
STEP 3: Similarly, we can find the x-intercept and the y-intercept of the graph of the
equation y = x – 2.
We get the x-intercept as (0, –2) and the y-intercept as (2, 0).
STEP 4: Mark the intercepts on the same coordinate plane and join them using a straight
line to get the graph of y = x – 2.
STEP 5: To obtain the solution, find the intersection point of the graphs.
The intersection point is (3, 1).
The solution of the system of equations is thus x = 3 and y = 1.
Example 2:
Solve the system of equations graphically and check your solution: x = y + 1 and 2x + y =
7
y = x -1
y = -2x - 7
y = x - 1 y = -2x - 7

slope = 1 slope = -2
y-intercept = -1 y-intercept = -7
INEQUALITY AND ITS PROPERTIES:
An inequality is a statement that one quantity or expression is greater or less than
another quantity or expression. Mathematically, for any two real numbers a and b, (a < b) or
(b > a) if and only if (b – a) is a positive number, that is, if and only if b-a > 0.
Inequality is the opposite of equality because equality (also called equation) states
that one quantity or expression is exactly equal to another quantity or expression.
Inequalities with <, > or ≠, are called strict inequalities.
Symbolically, x = y + 1 2x + y = -7 inequalities are

written and -2 = -3 + 1 2(-2) + (-3) = -7 read as follows:
-2 = -2 (check) -4 + (-3) = -7
1. x > y -7 = -7 (check) means “x is
greater than y”, (or x –
y is a positive number).
2. x < y means “x is less than y”, (or x – y is a negative number).
3. x ≥ y means “x is greater than or equal to y”.
4. x ≤ y means “x is less than or equal to y”.
5. x < y < z means “y is greater than x but less than z”.
6. x ≤ y < z means “y is greater than or equal to x but less than z”.
A. Trichotomy Property. For any real numbers a, and b, exactly one of the is true:
x < y; x=y; x > y
Example 1. If a = 3 and b = 5, only one of the three (<, = >) will hold true in any two given
numbers, thus, 3 < 5.
Example 2. Alex Has More Money Than Billy
We could write it like this: x>y
So, we also know that:
Alex does not have less money than Billy (not x<y)
Alex does not have the same amount of money as Billy (not x=y)
B. Addition and Subtraction Property

If x < y, then x ± z < y ± z
If x > y, then x ± z > y ± z
C. Mul
tipli
cati
on
Property
Note: If the multiplier is positive the inequality sign does not change, but if is negative the
inequality will be reverse.
D. Division Property
Note: If the divisor is positive the inequality sign does not change, but if is negative the
E. Exponential Property
If x < y and z is positive, then xz < yz.
If x < y and z is negative, then xz > yz.
Example: Given x = 2, y = 3, and z = 4
2 < 3 → 24 < 34 → 16 < 81
On the contrary, if x = 2, y = 3, and z = -4
1 1 1 1
2 < 3 → 2-4 >3-4 → > 4→ >
4
2 3 16 81
Note: If the divisor is positive the inequality sign does not change, but if is negative the
TYPES OF INEQUALITIES
A. Absolute Inequality. It is a condition where the inequality is true for all values of the
unknown involved.
Example: x2 + 4 > 0
It is an absolute inequality since every value of x (whether x is positive negative or zero) is
true for x2 > -4.
B. Conditional Inequality. It is a condition where the inequality is true for certain values
of the unknown involved.
Example 2x – 5 >7
2x – 5 + 5 > 7 + 5 Adding 5 to both sides.
2x > 7 + 5 Apply additive inverse properly.
2x > 12 Combine the terms.
( 12 ) ( 2 x ) >( 12 ) ( 12) Multiply both sides by ½.
x¿( )
12
Reducing to lowest term.
2
x ¿6 Solution
GRAPHING LINEAR INEQUALITY

In graphing linear inequalities, we will use the slope intercept and the test point in
identifying the solution set.
Steps in Graphing Inequalities
1. Replace the inequality sign with an equal sign, and then plot the graph of the
equation.
2. Select a test point lying in one of the half-planes determine by the graph and
substitute the values of x and y into the given inequality. Apply the origin test
whenever possible.
3. If the inequality is satisfied, the graph of the inequality includes the half-planes that
contain the test point. Otherwise, the solution includes the other half-plane
Example 1.
2y − x ≤ 6
Solution
To graph this inequality, start by making y the subject of the formula.
Adding x to both sides gives;
2y ≤ x + 6
Divide both sides by 2;
y ≤ x/2 + 3
Now plot the equation of y = x/2 + 3 as a solid line because of the ≤ sign. Shade below the
line because of the ≤ sign.
Example 2.
y/2 + 2 > x
Solution
Make y the subject of the formula.
Subtract both sides by 2;
y/2 > x − 2
Multiply both sides by 2 to eliminate the fraction:
y > 2x − 4
Now, because of the > sign, plot a dashed line of y = 2x − 4.
Example 3.
Solve the following inequality by graphing: 2x – 3y ≥ 6
Solution
The first is to make y the subject of the line 2x – 3y ≥ 6.
Subtract 2x from both sides of the equation.
2x – 2x – 3y ≥ 6 – 2x
-3y ≥ 6 – 2x
Divide both sides by -3 and reverse the sign.
y ≤ 2x/3 -2
Now draw a graph of y = 2x/3 – 2 and shade below the line.
Example 4.
x + y < 1
Solution
Rewrite the equation x + y = 1 so as to make y the subject of the formula. Because
the inequality sign is <, we will draw our graph with a dotted line.
After drawing the dotted line, we shade above the line because of the < sign.
Example 5.
Find the graphical solution of the following inequalities:
y≤x
y ≥ -x
x=5
Solution
Draw all the inequalities.
Red represents y ≤ x
Blue represents y ≥ -x
Green represents line x = 5
The common shaded area (can be seen clearly) is the graphical solution to these inequalities.
SYSTEM OF LINEAR INEQUALITIES

A system of linear inequalities is a set of equations of linear inequalities containing
the same variables. Several methods of solving systems of linear equations translate to the
system of linear inequalities. However, solving a system of linear inequalities is somewhat
different from linear equations because the inequality signs hinder us from solving by
substitution or elimination method. Perhaps the best method to solve systems of linear
inequalities is by graphing the inequalities.
How to Solve Systems of Linear Inequalities?

Previously, you learned how to solve a single linear inequality by graphing. In this
article, we will learn how to find solutions for a system of linear inequalities by graphing two
or more linear inequalities simultaneously.
The solution to a system of linear inequality is the region where the graphs of all linear
inequalities in the system overlap.
To solve a system of inequalities, graph each linear inequality in the system on the same
x-y axis by following the steps below:
Isolate the variable y in each linear inequality.
Draw and shade the area above the borderline using dashed and solid lines for the symbols
> and ≥ respectively.
Similarly, draw and shade the area below the borderline using dashed and solid lines for the
symbols < and ≤ respectively.
Shade the region where all the equations overlap or intersect. If there is no intersection
region, then we conclude the system of inequalities has no solution.
The solution set of linear inequalities is a plane, which is common or the intersection
of the inequalities.
Example 1
Graph the following system of linear inequalities:
y ≤ x – 1 and y < –2x + 1
Solution
Graph the first inequality y ≤ x − 1.
Because of the “less than or equal to” symbol, we will draw a solid border and do the
shading below the line.
Also, graph the second inequality y < –2x + 1 on the same x-y axis.
In this case, our borderline will be dashed or dotted because of the less than symbol. Shade
the area below the borderline.
Therefore, the solution to this system of inequalities is the darker shaded region extending
forever in a downward direction, as shown below.
Example 2
Solve the following system of inequalities:
x – 5y ≥ 6
3x + 2y > 1
Solution
First, isolate the variable y to the left in each inequality.
For x – 5y ≥ 6;
=> x ≥ 6 + 5y
=> 5y ≤ x – 6
=> y ≤ 0.2x – 1.2
And for 3x + 2y > 1;
=> 2y > 1 – 3x
=> y > 0.5 – 1.5x
We’ll graph y ≤ 2x– 1.2 and y > 0.5 – 1.5x using a solid line and a broken, respectively.
The solution of the system of inequality is the darker shaded area which is the overlap of the
two individual solution regions.
Example 3
Graph the following system of linear inequalities.
y ≤ (1/2) x + 1,
y ≥ 2x – 2,
y ≥ -(1/2) x – 3.
Solution
This system of inequalities has three equations that are all connected by an “equal to”
symbol. This tells us that all the borderlines will be solid. The graph of the three inequalities
is shown below.
The shaded region of the three equations overlaps right in the middle section. Therefore,
the solutions of the system lie within the bounded region, as shown on the graph.
POST-TEST:
Solve the following items:
1. Determine the type of system of linear equation using ratio test & plot the graphs.
a. 4x + 2y = 16
2x + y = 11
b. 2x + y = 6
4x + 2y = 12
c. 2x + y = 16
x + 2y = 11
2. Solve the system of linear equation of 2x + y = 16 and x + 2y = 11 using the five
methods: elimination, substitution, equivalent equation, Cramer’s rule and graphical.
3. Sketch the graph of 2x + y ≤ 16.
4. Find the solution set of the given inequalities using graphical method x + 2y ≥ 4 and
2x + y ≤ 6.
LESSON 2: LINEAR PROGRAMMING: GRAPHICAL METHOD

PRE-TEST:
1. What is Linear, Programming & Linear Programming?
2. What is the importance of Linear Programming?
3. How linear programming help the students in the study of Quantitative Business
Analysis?
1. Identify decision variables.
2. Formulate the objective function and the explicit constraints of the problem;
3. Determine the feasible solution area utilizing the graphical method of solving linear
programming problems; and
4. Formulate the decisions of linear programming problems if an optimal solution is
reached.
Every student in business administration has a secret dream of becoming an executive or

manager after graduation. A good manager can maximize the company’s profit and
minimize its cost without violating any limitation or restriction on existing variables, such as
time and quantity of available raw materials. Linear programming is a vital tool in
management science. This quantitative technique in management is so important that it
results in enormous savings of money.
Optimization is the way of life. We all have finite resources and time and we want to
make the most of them. From using your time productively to solving supply chain problems
for your company – everything uses optimization. It’s an especially interesting and relevant
topic in data science.
It is also a very interesting topic – it starts with simple problems, but it can get very
complex. For example, sharing a bar of chocolate between siblings is a simple optimization
problem. We don’t think in mathematical terms while solving it. On the other hand, devising
inventory and warehousing strategy for an e-tailer can be very complex.
Linear programming (LP) is one of the simplest ways to perform optimization. It
helps you solve some very complex optimization problems by making a few simplifying
assumptions. As an analyst, you are bound to come across applications and problems to be
solved by Linear Programming. For some reason, LP doesn’t get as much attention as it
deserves while learning data science.
Linear – implies direct proportionality of relationship of variables.
Programming – means making schedules or plans of activities to undertake in the
future.
Linear Programming – is the planning by the use of linear relationship of variables
involved. It makes use of certain mathematical techniques to get the best possible
solution to a problem involving limited resources.
A minimization or maximization problem can be classified as linear programming if it
has the following properties:
1. The objective of the decision-maker must be either to maximize or to minimize.
2. The actions of the decision-maker must be constrained and the decision variables must
not violate the limitations or constraints.
3. All variables have a value greater than or equal to 0.
4. The relationship of variables could be expressed in terms of equations or inequalities.
 Solving Linear Programming Problem Graphically

A. Linear Programming Problems
A linear programming problem in two unknowns x and y is one in which we are to

determine the maximum and minimum value of a linear expression.
P = a1x + b1y (for maximization) C = a1x + b1y (for minimization)
called the objective function, subject to a number of linear constraints of the form
a1x + b1y ≤ c1 or a1x + b1y ≥ c1 or a1x + b1y = c1
An objective function is an expression, which shows the relationship between the
variables in the problem and the firm’s goal.
Two types of constraints:
a. Structural constraint – is a limit on the availability of resources; it is also referred as

explicit constraints.
b. Negativity constraint – is the constraint that restricts all the variables to zero and
positive solution; it is also referred as implicit constraint.
The Linear programming model:

Maximize: P = 1,200x + 1,600y Objective Function
Subject to: 3x + 2y ≤ 18
2x + 4y ≤ 20 Structural Constraints
x ≤5
x ≥ 0, y ≥ 0 Non-negativity Constraints
Optimal value (for maximization problem) or lowest value (for minimization problem)
of the objective function.
Optimal solution – is a combination of decision variable amounts that yield the best
possible value of the objective function and satisfies all the constraints.
Feasible region – is the set of combinations of values for the decision variables that
satisfy the non-negativity conditions and all the constraints simultaneously that is the
allowable decisions.
Extreme point – is the corner of the feasible region; it is the location of the maximum
and minimum point of the feasible region.
B. The Extreme Point Theorem
The linear objective function will have its optimal solutions at the extreme points
(corner points) of the feasible region whenever the feasible region is bounded. If the
feasible region is unbounded, there is no optimal solution. In cases wherein there is an
optimal solution even though the feasible region is unbounded, it lies at the extreme (or
corner) of the feasible region.
C. Fundamental Theorem of Linear Programming Problem
There are two things to consider in solving linear programming problem:
 If a Linear Programming (LP) problem has optimal solution, there is always at least
one extreme point (corner point) solution of the feasible region.
 A Linear Programming (LP) problem with bounded, nonempty feasible regions
always contain optimal solutions.
D. Solving Linear Programming Maximization Problem

 Steps in Linear Programming Graphical Method
1. Graph the linear inequalities and determine the feasible region.
2. Determine the coordinates of the extreme points (corner points).
3. Substitute the coordinates of the extreme points to the objective function and
identify the highest (for maximization problem) or lowest (for minimization problem)
result.
4. The highest (for maximization problem) or lowest minimization problem) result
obtained in step 3 serves as the optimal solution to the LP problem.
Example 1.) The demand for two products in each of the last four weeks is shown below.
Week
1234
Demand - product 1 23 27 34 40
Demand - product 2 11 13 15 14
Apply exponential smoothing with a smoothing constant of 0.7 to generate a

forecast for the demand for these products in week 5.
These products are produced using two machines, X and Y. Each unit of product 1 that is
produced requires 15 minutes processing on machine X- and 25-minutes processing on
machine Y. Each unit of product 2 that is produced requires 7 minutes processing on
machine X- and 45-minutes processing on machine Y. The available time on machine X in
week 5 is forecast to be 20 hours and on machine Y in week 5 is forecast to be 15 hours.
Each unit of product 1 sold in week 5 gives a contribution to profit of £10 and each unit of
product 2 sold in week 5 gives a contribution to profit of £4.
It may not be possible to produce enough to meet your forecast demand for these products
in week 5 and each unit of unsatisfied demand for product 1 costs £3, each unit of
unsatisfied demand for product 2 costs £1.
Formulate the problem of deciding how much of each product to make in week 5 as
a linear program.
Solve this linear program graphically.
Solution
Note that the first part of the question is a forecasting question so it is solved below.
For product 1 applying exponential smoothing with a smoothing constant of 0.7 we get:
M1 = Y1 = 23
M2 = 0.7Y2 + 0.3M1 = 0.7(27) + 0.3(23) = 25.80
M3 = 0.7Y3 + 0.3M2 = 0.7(34) + 0.3(25.80) = 31.54
M4 = 0.7Y4 + 0.3M3 = 0.7(40) + 0.3(31.54) = 37.46
The forecast for week five is just the average for week 4 = M 4 = 37.46 = 31 (as we cannot
have fractional demand).
For product 2 applying exponential smoothing with a smoothing constant of 0.7 we get:
M1 = Y1 = 11
M2 = 0.7Y2 + 0.3M1 = 0.7(13) + 0.3(11) = 12.40
M3 = 0.7Y3 + 0.3M2 = 0.7(15) + 0.3(12.40) = 14.22
M4 = 0.7Y4 + 0.3M3 = 0.7(14) + 0.3(14.22) = 14.07
The forecast for week five is just the average for week 4 = M 4 = 14.07 = 14 (as we cannot
have fractional demand).
We can now formulate the LP for week 5 using the two demand figures (37 for product 1
and 14 for product 2) derived above.
Let
x1 be the number of units of product 1 produced
x2 be the number of units of product 2 produced
where x1, x2>=0
The constraints are:
15x1 + 7x2 <= 20(60) machine X
25x1 + 45x2 <= 15(60) machine Y
x1 <= 37 demand for product 1
x2 <= 14 demand for product 2
The objective is to maximize profit, i.e.
maximize 10x1 + 4x2 - 3(37- x1) - 1(14-x2)
i.e. maximize 13x1 + 5x2 - 125
The graph is shown below, from the graph we have that the solution occurs on the
horizontal axis (x2=0) at x1=36 at which point the maximum profit is 13(36) + 5(0) - 125 =
£343
Example 2.) A company makes two products (X and Y) using two machines (A and B). Each
unit of X that is produced requires 50 minutes processing time on machine A and 30 minutes
processing time on machine B. Each unit of Y that is produced requires 24 minutes
processing time on machine A and 33 minutes processing time on machine B.
At the start of the current week there are 30 units of X and 90 units of Y in stock. Available
processing time on machine A is forecast to be 40 hours and on machine B is forecast to be
35 hours.
The demand for X in the current week is forecast to be 75 units and for Y is forecast
to be 95 units. Company policy is to maximize the combined sum of the units of X and the
units of Y in stock at the end of the week.
Formulate the problem of deciding how much of each product to make in the
current week as a linear program.
Solve this linear program graphically.
Solution
Let:
x be the number of units of X produced in the current week
y be the number of units of Y produced in the current week
then the constraints are:
50x + 24y <= 40(60) machine A time
30x + 33y <= 35(60) machine B time
x >= 75 - 30
i.e. x >= 45 so production of X >= demand (75) - initial stock (30), which ensures we
meet demand
y >= 95 - 90
i.e. y >= 5 so production of Y >= demand (95) - initial stock (90), which ensures we
meet demand
The objective is: maximize (x+30-75) + (y+90-95) = (x+y-50)
i.e. to maximize the number of units left in stock at the end of the week
It is plain from the diagram below that the maximum occurs at the intersection of x=45 and
50x + 24y = 2400
Solving
simultaneously, rather than by reading values off the graph, we have that x=45 and y=6.25
with the value of the objective function being 1.25
Example 3.)
E. Solving Linear Programming Minimization Problem
The same procedure in establishing the solution of an LP with maximization problem the
only difference is select the lowest linear combination in the objective function in contrast
with the maximization problem.
Example 1.)
Example
2.)
Example 3.) Solve
Minimize: 4a + 5b + 6c
subject to: a + b >= 11
a - b <= 5
c-a-b=0
7a >= 35 - 12b
a >= 0 b >= 0 c >= 0
Solution
To solve this LP, we use the equation c-a-b=0 to put c=a+b (>= 0 as a >= 0 and b >= 0) and so
the LP is reduced to
Minimize: 4a + 5b + 6(a + b) = 10a + 11b
subject to: a + b >= 11
a - b <= 5
7a + 12b >= 35
a >= 0 b >= 0
From the diagram below the minimum occurs at the intersection of a - b = 5 and a + b = 11
i.e. a = 8 and b = 3 with c (= a + b) = 11 and the value of the objective function 10a + 11b =
80 + 33 = 113.
 Linear Programming Problem Involving Mixed Constraints

A. Maximization Involving Mixed Constraint
Example 2.)
Only change the objective function into minimization problem instead. The solution will
obtain the same result, the only difference is that we must choose the lowest linear
combination in the optimal solution.
POST-TEST:
Solve the following:
1.) A local boutique produced two designs of gowns A and B and has the following materials
available: 18 square meters of cotton, 20 square meters of silk, and 5 square meters of wool.
Design A requires the following: 3 square meters of cotton, 2 square meters of silk and 1
square meter of wool. Design B requires the following: 2 square meters of cotton, 4 square
meters of silk. If Design A sells for P1,200 and Design B for P1,600, how many of each
garment should the boutique produce to obtain the maximum amount of money?
2.) A pharmacist produces a drug from two ingredients. Each ingredient contains the same
three antibiotics in different proportions. Each ingredient A produced results P80 in cost;
each ingredient B results P50 in cost. The production of the antibiotics is dependent on the
availability of limited resources. The resource requirements for the production are as
follows:
Resources Requirement
Antibiotic Ingredient A Ingredient B Minimum Requirement
Antibiotic 1 3 units 1 unit 6
Antibiotic 2 1 unit 1 unit 4
Antibiotic 3 2 units 6 units 12
The company wants to determine the quantity of ingredient A and B that must go in to drug
in order to meet the antibiotics minimum requirements at the minimum cost.
3.) Solve the linear programming problem:

Maximize: P = 6x + 4y
Subject to: 3x + 2y ≥ 18
2x + 4y = 20
2y ≤ 8
x ≥ 0, y ≥ 0
LESSON 3: LINEAR PROGRAMMING: THE SIMPLEX METHOD
PRE-TEST:
1. Discuss the difference between the simplex method and the graphical method of
linear programming.
2. Enumerate the procedure for the simplex maximation method.
3. Enumerate the procedure for the simplex minimization method.
1. Investigate real world applications of linear programming and related methods.
2. Differentiate the simplex method from the graphical method of linear programming.
3. Solve linear programming maximization problems using the simplex method.
4. Solve linear programming minimization problems using the simplex method.
5. Identify and set up a linear program in standard minimization form
6. Formulate a dual problem in standard maximization form
7. Use the simplex method to solve the dual maximization problem
8. Identify the optimal solution to the original minimization problem from the optimal
simplex tableau.
Most real-world linear programming problems have more than two variables and thus
are too complex for graphical solution. A procedure called the simplex method may be used
to find the optimal solution to multivariable problems. The simplex method is actually an
algorithm (or a set of instructions) with which we examine corner points in a methodical
fashion until we arrive at the best solution—highest profit or lowest cost. Computer
programs and spreadsheets are available to handle the simplex calculations for you. But you
need to know what is involved behind the scenes in order to best understand their valuable
outputs.
What is maximization in linear programming?
The Fundamental Theorem of Linear Programming states that the maximum (or
minimum) value of the objective function always takes place at the vertices of the feasibility
region. ... To maximize income, we will substitute these points in the objective function to
see which point gives us the highest income per week.
The Simplex Method: Standard Maximation Problems
Solving linearly programming problems graphically is ideal, but with large numbers
of constraints or variables, doing so becomes unreasonable. A tool to solve these problems,
is the Simplex Method, as a black-box tool that can only be used to solve standard
maximization problems.
Example
Niki holds two part-time jobs, Job I and Job II. She never wants to work more than a
total of 12 hours a week. She has determined that for every hour she works at Job I, she
needs 2 hours of preparation time, and for every hour she works at Job II, she needs one
hour of preparation time, and she cannot spend more than 16 hours for preparation. If she
makes $40 an hour at Job I, and $30 an hour at Job II, how many hours should she work per
week at each job to maximize her income?
Solution
STEP 1. Set up the problem. Write the objective function and the constraints.
Since the simplex method is used for problems that consist of many variables, it is not
practical to use the variables x, y, z etc. We use symbols x1, x2, x3, and so on.
Let
 x1 = The number of hours per week Niki will work at Job I and
 x2 = The number of hours per week Niki will work at Job II.
It is customary to choose the variable that is to be maximized as Z.
The problem is formulated the same way as we did in the last lesson.
Maximize Z=40x1+30x2
Subject to: x1+x2≤12
2x1+x2≤16
x1≥0; x2≥0
STEP 2. Convert the inequalities into equations. This is done by adding one slack variable
for each inequality.
For example, to convert the inequality x1+x2≤12 into an equation, we add a non-negative
variable y1, and we get
x1+x2+y1=12
Here the variable y1 picks up the slack, and it represents the amount by which x1+x2 falls
short of 12. In this problem, if Niki works fewer than 12 hours, say 10, then y1 is 2. Later
when we read off the final solution from the simplex table, the values of the slack variables
will identify the unused amounts.
We rewrite the objective function Z=40x1+30x2 as −40x1−30x2+Z=0.
After adding the slack variables, our problem reads
Objective function −40x1−30x2+Z=0
Subject to constraints: x1+x2+y1=12
2x1+x2+y2=16
x1≥0; x2≥0
STEP 3. Construct the initial simplex tableau. Each inequality constraint appears in its own
row. (The non-negativity constraints do not appear as rows in the simplex tableau.) Write
the objective function as the bottom row.
Now that the inequalities are converted into equations, we can represent the problem into
an augmented matrix called the initial simplex tableau as follows.
Here the vertical line separates the left-hand side of the equations from the right side. The
horizontal line separates the constraints from the objective function. The right side of the
equation is represented by the column C.
The reader needs to observe that the last four columns of this matrix look like the final
matrix for the solution of a system of equations. If we arbitrarily choose x1=0 and x2=0, we
get
which reads
y1=12 y2=16 Z=0
The solution obtained by arbitrarily assigning values to some variables and then solving for
the remaining variables is called the basic solution associated with the tableau. So, the
above solution is the basic solution associated with the initial simplex tableau. We can label
the basic solution variable in the right of the last column as shown in the table below.
STEP 4. The most negative entry in the bottom row identifies the pivot column.
The most negative entry in the bottom row is -40; therefore, the column 1 is identified.
Question Why do we choose the most negative entry in the bottom row?
Answer The most negative entry in the bottom row represents the largest coefficient in the
objective function - the coefficient whose entry will increase the value of the objective
function the quickest. The simplex method begins at a corner point where all the main
variables, the variables that have symbols such as x1, x2, x3 etc., are zero. It then moves from
a corner point to the adjacent corner point always increasing the value of the objective
function. In the case of the objective function Z=40x1+30x2, it will make more sense to
increase the value of x1 rather than x2. The variable x1 represents the number of hours per
week Niki works at Job I. Since Job I pays $40 per hour as opposed to Job II which pays only
$30, the variable x1 will increase the objective function by $40 for a unit of increase in the
variable x1.
STEP 5. Calculate the quotients. The smallest quotient identifies a row. The element in the
intersection of the column identified in step 4 and the row identified in this step is
identified as the pivot element.
Following the algorithm, in order to calculate the quotient, we divide the entries in
the far-right column by the entries in column 1, excluding the entry in the bottom row.
The smallest of the two quotients, 12 and 8, is 8. Therefore row 2 is identified. The
intersection of column 1 and row 2 is the entry 2, which has been highlighted. This is our
pivot element.
Question Why do we find quotients, and why does the smallest quotient identify a row?
Answer When we choose the most negative entry in the bottom row, we are trying to
increase the value of the objective function by bringing in the variable x1. But we cannot
choose any value for x1. Can we let x1=100? Definitely not! That is because Niki never wants
to work for more than 12 hours at both jobs combined: x1+x2≤12. Can we let x1=12? Again,
the answer is no because the preparation time for Job I is two times the time spent on the
job. Since Niki never wants to spend more than 16 hours for preparation, the maximum
time she can work is 16 ÷ 2 = 8. Now you see the purpose of computing the quotients; using
the quotients to identify the pivot element guarantees that we do not violate the
constraints.
Question Why do we identify the pivot element?
Answer As we have mentioned earlier, the simplex method begins with a corner point and
then moves to the next corner point always improving the value of the objective function.
The value of the objective function is improved by changing the number of units of the
variables. We may add the number of units of one variable, while throwing away the units
of another. Pivoting allows us to do just that.
The variable whose units are being added is called the entering variable, and the variable
whose units are being replaced is called the departing variable. The entering variable in the
above table is x1, and it was identified by the most negative entry in the bottom row. The
departing variable y2 was identified by the lowest of all quotients.
STEP 6. Perform pivoting to make all other entries in this column zero.
In chapter 2, we used pivoting to obtain the row echelon form of an augmented matrix.
Pivoting is a process of obtaining a 1 in the location of the pivot element, and then making
all other entries zeros in that column. So now our job is to make our pivot element a 1 by
dividing the entire second row by 2. The result follows.
To obtain a zero in the entry first above the

pivot element, we multiply the second row by -1 and add it to row 1. We get
To obtain a zero in the element below the pivot,

we multiply the second row by 40 and add it to
the last row.
We now determine the basic solution associated with this tableau. By arbitrarily
choosing x2=0 and y2=0, we obtain x1=8, y1=4, and z=320. If we write the augmented matrix,
whose left side is a matrix with columns that have one 1 and all other entries zeros, we get
the following matrix stating the same thing.
We can restate the solution associated with this matrix as x1=8, x2=0, y1=4, y2=0 and z=320.

At this stage of the game, it reads that if Niki works 8 hours at Job I, and no hours at Job II,
her profit Z will be $320. Recall from Example 3.1.1 in section 3.1 that (8, 0) was one of our
corner points. Here y1=4 and y2=0 mean that she will be left with 4 hours of working time
and no preparation time.
STEP 7. When there are no more negative entries in the bottom row, we are finished;
otherwise, we start again from step 4.
Since there is still a negative entry, -10, in the bottom row, we need to begin, again, from
step 4. This time we will not repeat the details of every step, instead, we will identify the
column and row that give us the pivot element, and highlight the pivot element. The result
is as follows.
We make the pivot element 1 by multiplying row 1 by 2, and we get

Now to make all other entries as zeros in this column, we first multiply row 1 by -1/2 and
add it to row 2, and then multiply row 1 by 10 and add it to the bottom row.
We no longer have negative entries in the

bottom row; therefore, we are finished.
Question Why are we finished when there are no negative entries in the bottom row?
Answer The answer lies in the bottom row. The bottom row corresponds to the equation:
0x1+0x2+20y1+10y2+Z=400 or
z=400−20y1−10y2
Since all variables are non-negative, the highest value Z can ever achieve is 400, and that will
happen only when y1 and y2 are zero.
STEP 8. Read off your answers.
We now read off our answers, that is, we determine the basic solution associated with the
final simplex tableau. Again, we look at the columns that have a 1 and all other entries
zeros. Since the columns labeled y1 and y2 are not such columns, we arbitrarily choose y1=0,
and y2=0, and we get
The matrix reads x1=4, x2=8 and z=400.

The final solution says that if Niki works 4 hours at Job I and 8 hours at Job II, she will
maximize her income to $400. Since both slack variables are zero, it means that she would
have used up all the working time, as well as the preparation time, and none will be left.
Simplex Solution of a Minimization Problem
Minimization linear programming problems are solved in much the same way as the
maximization problems. For the standard minimization linear program, the constraints are
of the form ax+by≥c, as opposed to the form ax+by≤c for the standard maximization
problem.
In the previous lesson the simplex method for solving linear programming problems
was demonstrated for a maximization problem. In general, the steps of the simplex method
outlined are used for any type of linear programming problem. However, a minimization
problem requires a few changes in the normal simplex process.
In addition, several exceptions to the typical linear programming problem will be
presented later in this module. These include problems with mixed constraints (=, ,
and ); problems with more than one optimal solution, no feasible solution, or an
unbounded solution; problems with a tie for the pivot column; problems with a tie for the
pivot row; and problems with constraints with negative quantity values. None of these kinds
of problems require changes in the simplex method. They are basically unusual results in
individual simplex tables that the reader should know how to interpret and work with.
Standard Form of a Minimization Model
Consider the following linear programming model for a farmer purchasing fertilizer.
where:
x1 = bags of Super-gro fertilizer
x2 = bags of Crop-quick fertilizer
Z = farmer's total cost ($) of purchasing fertilizer
This model is transformed into standard form by subtracting surplus variables from the
two constraints as follows.
The surplus variables represent the extra amount of nitrogen and phosphate that exceeded
the minimum requirements specified in the constraints.
However, the simplex method requires that the initial basic feasible solution be at the
origin, where x1 = 0 and x2 = 0. Testing these solution values, we have
2 x1 + 4 x2 S1 = 16
2(0) +4(0) S1 = 16
S1 = 16
The idea of “negative excess pounds of nitrogen" is illogical and violates the nonnegativity
restriction of linear programming. The reason the surplus variable does not work is shown in
Figure A-4. The solution at the origin is outside the feasible solution space.
Figure A-4. Graph of the fertilizer example
Transforming a model into standard form by subtracting surplus variables will not
work in the simplex method.
To alleviate this difficulty and get a solution at the origin, we add an artificial variable (A1 ) to
the constraint equation,
2x1 + 4x2 S1 + A1 = 16
2x1 + 4x2 S1 + A1 = 16
An artificial variable allows for an initial basic
2(0) + 4(0) 0 + A1 = 16
feasible solution at the origin, but it has no real meaning.
A1 = 16
The artificial variable, A1, does not have a
meaning as a slack variable or a surplus variable does. It is inserted into the equation simply
to give a positive solution at the origin; we are artificially creating a solution.
The artificial variable is somewhat analogous to a booster rocket its purpose is to get
us off the ground; but once we get started, it has no real use and thus is discarded. The
artificial solution helps get the simplex process started, but we do not want it to end up in
the optimal solution, because it has no real meaning.
When a surplus variable is subtracted and an artificial variable is added, the
phosphate constraint becomes
4x1 + 3x2 S2 + A2 = 24
The effect of surplus and artificial variables on the objective function must now be
considered. Like a slack variable, a surplus variable has no effect on the objective function in
terms of increasing or decreasing cost. For example, a surplus of 24 pounds of nitrogen does
not contribute to the cost of the objective function, because the cost is determined solely by
the number of bags of fertilizer purchased (i.e., the values of x1 and x2 ). Thus, a coefficient of
0 is assigned to each surplus variable in the objective function.
By assigning a "cost" of $0 to each surplus variable, we are not prohibiting it from
being in the final optimal solution. It would be quite realistic to have a final solution that
showed some surplus nitrogen or phosphate. Likewise, assigning a cost of $0 to an artificial
variable in the objective function would not prohibit it from being in the final optimal
solution. However, if the artificial variable appeared in the solution, it would render the final
solution meaningless. Therefore, we must ensure that an artificial variable is not in the final
solution.
As previously noted, the presence of a particular variable in the final solution is
based on its relative profit or cost. For example, if a bag of Super-gro costs $600 instead of
$6 and Crop-quick stayed at $3, it is doubtful that the farmer would purchase Super-gro
(i.e., x 1 would not be in the solution). Thus, we can prohibit a variable from being in the final
solution by assigning it a very large cost. Rather than assigning a dollar cost to an artificial
variable, we will assign a value of M, which represents a large positive cost (say,
$1,000,000). This operation produces the following objective function for our example:
minimize Z = 6x1 + 3x2 + 0S1 + 0S2 + MA1 + MA2
The completely transformed minimization model can now be summarized as
Artificial variables are assigned a large cost in the objective function to eliminate them from
the final solution.
The Simplex Table for a Minimization Problem
The initial simplex tableau for a minimization model is developed the same way as
one for a maximization model, except for one small difference. Rather than
computing c j z j in the bottom row of the tableau, we compute z j c j , which represents
the net per unit decrease in cost , and the largest positive value is selected as the entering
variable and pivot column. (An alternative would be to leave the bottom row as c j z j and
select the largest negative value as the pivot column. However, to maintain a consistent rule
for selecting the pivot column, we will use z j c j .)
The c j z j row is changed to z j c j in the simplex table for a minimization problem.
The initial simplex tableau for this model is shown in Table A-17. Notice
that A 1 and A 2 form the initial solution at the origin, because that was the reason for
inserting them in the first place to get a solution at the origin. This is not a basic feasible
solution, since the origin is not in the feasible solution area, as shown in Figure A-4. As
indicated previously, it is an artificially created solution. However, the simplex process will
move toward feasibility in subsequent tableaus. Note that across the top the decision
variables are listed first, then surplus variables, and finally artificial variables
Artificial variables are always included as part of the initial basic feasible solution when they
exist.
In Table A-17 the x 2 column was selected as the pivot column because 7 M 3 is the
largest positive value in the z j c j row. A 1 was selected as the leaving basic variable (and pivot
row) because the quotient of 4 for this row was the minimum positive row value.
Once an artificial variable is selected as the leaving variable, it will never reenter the
tableau, so it can be eliminated.
The second simplex table is developed using the simplex formulas presented earlier.
It is shown in Table A-18. Notice that the A1 column has been eliminated in the second
simplex table. Once an artificial variable leaves the basic feasible solution, it will never
return because of its high cost, M. Thus, like the booster rocket, it can be eliminated from
the tableau. However, artificial variables are the only variables that can be treated this way.
The third simplex table, with x 1 replacing A2, is shown in Table A-19. Both
the A 1 and A 2 columns have been eliminated because both variables have left the solution.
The x 1 row is selected as the pivot row because it corresponds to the minimum positive
ratio of 16. In selecting the pivot row, the 4 value for the x 2 row was not considered because
the minimum positive value or zero is selected. Selecting the x 2 row would result in a
negative quantity value for s 1 in the fourth tableau, which is not feasible.
The fourth simplex table, with s 1 replacing x1, is shown in Table A-20. Table A-20 is
the optimal simplex tableau because the z j c j row contains no positive values. The optimal
solution is
x1 = 0 bags of Super-gro
S1 = 16 extra lb of nitrogen
x2 = 8 bags of Crop-quick
S2 = 0 extra lb of phosphate
Z = $24, total cost of purchasing fertilizer
Simplex Adjustments for a Minimization Problem
To summarize, the adjustments necessary to apply the simplex method to a
minimization problem are as follows:
Transform all constraints to equations by subtracting a surplus variable and
adding an artificial variable.
Change the cj zj row to zj cj.
Although the fertilizer example model, we just used to include only constraints, it
is possible for a minimization problem to have and = constraints in addition to
constraints. Similarly, it is possible for a maximization problem to have and = constraints
in addition to constraints. Problems that contain a combination of different types of
inequality constraints are referred to as mixed constraint problems.
The Duality Principle
The objective function of the minimization problem reaches its minimum if and only
if the objective function of its dual reaches its maximum. And when they do, they are equal.
Our next goal is to extract the solution for our minimization problem from the
corresponding dual. To do this, we solve the dual by the simplex method.
Example
Find the solution to the minimization problem by solving its dual using the simplex method.
We rewrite our problem.
Solution
Recall that we solved the above problem by the simplex method. Therefore, we only show
the initial and final simplex tableau.
The initial simplex table is
Observe an important change. Here our main variables are y1 and y2 and the slack variables
are x1 and x2.
The final simplex table reads as follows:
A closer look at this table reveals that the x1 and x2 values along with the minimum
value for the minimization problem can be obtained from the last row of the final tableau.
We have highlighted these values by the arrows.
We restate the solution as follows:

The minimization problem has a minimum value of 400 at the corner point (20, 10).
We now summarize our discussion.
Minimization by Simplex Method
1. Set up the problem.
2. Write a matrix whose rows represent each constraint with the objective function as
its bottom row.
3. Write the transpose of this matrix by interchanging the rows and columns.
4. Now write the dual problem associated with the transpose.
5. Solve the dual problem by the simplex method.
6. The optimal solution is found in the bottom row of the final matrix in the columns
corresponding to the slack variables, and the minimum value of the objective
function is the same as the maximum value of the dual.
POST-TEST:
Solve the following items:
1. Maximize P = 40x1 + 30x2
Subject to: x1 + 2x2 ≤ 16
x1 + x2 ≤ 9
3x1 + 2x2 ≤ 24
x1 , x2 ≥ 0
2.
3. Find solution using Simplex (Big M) method
MIN Z = x 1 + x2
subject to
2x1 + 4x2 >= 4
x1 + 7x2 >= 7
and x1, x2 >= 0
4. A small petroleum company owns two refineries. Refinery 1 costs $20,000 per day to
operate, and it can produce 400 barrels of high-grade oil, 300 barrels of medium-
grade oil, and 200 barrels of low-grade oil each day. Refinery 2 is newer and more
modern. It costs $25,000 per day to operate, and it can produce 300 barrels of high-
grade oil, 400 barrels of medium-grade oil, and 500 barrels of low-grade oil each day.
The company has orders totaling 25,000 barrels of high-grade oil, 27,000 barrels of
medium-grade oil, and 30,000 barrels of low-grade oil. How many days should it run
each refinery to minimize its costs and still refine enough oil to meet its orders?
5.
LESSON 4: SPECIAL PURPOSE ALGORITHMS PROGRAMMING

(TRANSPORTATION & ASSIGNMENT MODELS)
PRE-TEST:
1. Discuss Transportation Models.
2. Discuss Assignment Models.
3. Explain the uses of transportation & Assignment Models.
1. Identify the special features of the transportation problem.
2. Familiar with the types of problems that can be solved by applying a transportation
model.
3. Develop network and linear programming models of the transportation problem.
4. Identify the special features of the assignment problem.
5. Familiar with the types of problems that can be solved by applying an assignment
model.
6. Develop network and linear programming models of the assignment problem.
7. Familiar with the special features of the transshipment problem.
8. Familiar with the types of problems that can be solved by applying a transshipment
model.
Transportation and Assignment Models

Transportation and assignment models are special purpose algorithms of the linear
programming. The simplex method of Linear Programming Problems (LPP) proves to be
inefficient is certain situations like determining optimum assignment of jobs to persons,
supply of materials from several supply points to several destinations and the like. More
effective solution models have been evolved and these are called assignment and
transportation models.
The transportation model is concerned with selecting the routes between supply and
demand points in order to minimize costs of transportation subject to constraints of supply
at any supply point and demand at any demand point. Assume a company has 4
manufacturing plants with different capacity levels, and 5 regional distribution centers, 4 x 5
= 20 routes are possible. Given the transportation costs per load of each of 20 routes
between the manufacturing (supply) plants and the regional distribution (demand) centers,
and supply and demand constraints, how many loads can be transported through different
routes so as to minimize transportation costs? The answer to this question is obtained
easily through the transportation algorithm.
Similarly, how are we to assign different jobs to different persons/machines, given
cost of job completion for each pair of job machine/person? The objective is minimizing
total cost. This is best solved through assignment algorithm.
Uses of Transportation and Assignment Models in Decision Making
Transportation model is used in the following:
 To decide the transportation of new materials from various centers to different
manufacturing plants. In the case of multi-plant company this is highly useful.
 To decide the transportation of finished goods from different manufacturing plants
to the different distribution centers. For a multi-plant-multi-market company this is
useful.
 To decide the transportation of finished goods from different manufacturing plants
to the different distribution centers. For a multi-plant-multi-market company this is
useful. These two are the uses of transportation model. The objective is
minimizing transportation cost.
Assignment model is used in the following:
 To decide the assignment of jobs to persons/machines, the assignment model is
used.
 To decide the route a traveling executive has to adopt (dealing with the order in
which he/she has to visit different places).
 To decide the order in which different activities performed on one and the same
facility be taken up.
In the case of transportation model, the supply quantity may be less or more than the
demand. Similarly, the assignment model, the number of jobs may be equal to, less or
more than the number of machines/persons available. In all these cases the simplex
method of LPP can be adopted, but transportation and assignment models are more
effective, less time consuming and easier than the LPP.
Transportation problem is a special kind of Linear Programming Problem (LPP) in which
goods are transported from a set of sources to a set of destinations subject to the supply
and demand of the sources and destination respectively such that the total cost of
transportation is minimized. It is also sometimes called as Hitchcock problem.
Types of Transportation problems:
1. Balanced: When both supplies and demands are equal then the problem is said to be
a balanced transportation problem.
2. Unbalanced: When the supply and demand are not equal then it is said to be an
unbalanced transportation problem. In this type of problem, either a dummy row or
a dummy column is added according to the requirement to make it a balanced
problem. Then it can be solved similar to the balanced problem.
Methods to Solve:
To find the initial basic feasible solution there are three methods:
1. Northwest Corner Cell Method.
2. Least Call Cell Method.
3. Vogel’s Approximation Method (VAM).
Basic structure of transportation problem:

In the above table D1, D2, D3 and D4 are the destinations where the products/goods
are to be delivered from different sources S1, S2, S3 and S4. Si is the supply from the source
Oi. dj is the demand of the destination Dj. Cij is the cost when the product is delivered from
source Si to destination Dj.
a) Transportation Problem: (Northwest Corner Method)
An introduction to Transportation problem has been discussed in the previous
Section, in this, finding the initial basic feasible solution using the Northwest Corner Cell
Method will be discussed.
Explanation: Given three sources O1, O2 and O3 and four destinations D1, D2, D3
and D4. For the sources O1, O2 and O3, the supply is 300, 400 and 500 respectively. The
destinations D1, D2, D3 and D4 have demands 250, 350, 400 and 200 respectively. Solution:
According to North West Corner method, (O1, D1) has to be the starting point i.e. the north-
west corner of the table. Each and every value in the cell is considered as the cost per
transportation. Compare the demand for column D1 and supply from the source O1 and
allocate the minimum of two to the cell (O1, D1) as shown in the figure.
The demand for Column D1 is completed so the entire column D1 will be canceled.
The supply from the source O1 remains 300 – 250 = 50.
Now from the remaining table i.e. excluding column D1, check the north-west corner
i.e. (O1, D2) and allocate the minimum among the supply for the respective column and the
rows. The supply from O1 is 50 which is less than the demand for D2 (i.e. 350), so allocate
50 to the cell (O1, D2).
Since the supply from row O1 is completed cancel the row O1. The demand for
column D2 remains 350 – 50 = 50.
From
the remaining table the north-west corner cell is (O2, D2). The minimum among the supply
from source O2 (i.e 400) and demand for column D2 (i.e 300) is 300, so allocate 300 to the
cell (O2, D2). The demand for the column D2 is completed so cancel the column and the
remaining supply from source O2 is 400 – 300 = 100.
Now from remaining table find the north-west corner i.e. (O2, D3) and compare the
O2supply (i.e. 100) and the demand for D2 (i.e. 400) and allocate the smaller (i.e. 100) to the
cell (O2, D2). The supply from O2 is completed so cancel the row O2. The remaining demand
for column D3 remains 400 – 100 = 300.
Note: In the last remaining cell the demand for the respective columns and rows are
equal which was cell (O3, D4). In this case, the supply from O3 and the demand for D4 was
200which was allocated to this cell. At last, nothing remained for any row or column.
Now just multiply the allocated value with the respective cell value (i.e. the cost) and add all
of them to get the basic solution i.e. (250 * 3) + (50 * 1) + (300 * 6) + (100 * 5) + (300 * 3) +
(200 * 2) = 4400.
b) Transportation Problem: (Least Cost Cell Method)
The North-West Corner method has been discussed in the previous session. In this
session, the Least Cost Cell method will be discussed.
Solution: According to the Least Cost Cell method, the least cost among all the cells in the
table has to be found which is 1 (i.e. cell (O1, D2)).
Now check the supply from the row O1 and demand for column D2 and allocate the
smaller value to the cell. The smaller value is 300 so allocate this to the cell. The supply from
O1 is completed so cancel this row and the remaining demand for the column D2 is 350 –
300 = 50.
Now find the cell with the least cost among the remaining cells. There are two cells
with the least cost i.e. (O2, D1) and (O3, D4) with cost 2. Let’s select (O2, D1). Now find the
demand and supply for the respective cell and allocate the minimum among them to the cell
and cancel the row or column whose supply or demand becomes 0 after allocation.
Now the cell with the least cost is (O3, D4) with cost 2. Allocate this cell with 200 as
the demand is smaller than the supply. So, the column gets canceled.
There are two cells among the unallocated cells that have the least cost. Choose any
at random say (O3, D2). Allocate this cell with a minimum among the supply from the
respective row and the demand of the respective column. Cancel the row or column with
zero value.
Now the cell with the least cost is (O3, D3). Allocate the minimum of supply and
demand and cancel the row or column with zero value.
The only remaining cell is (O2, D3) with cost 5 and its supply is 150 and demand is
150 i.e. demand and supply both are equal. Allocate it to this cell.
Now just multiply the cost of the cell with their respective allocated values and add all of
them to get the basic solution i.e. (300 * 1) + (25 * 2) + (150 * 5) + (50 * 3) + (250 * 3) + (200
* 2) = 2400
c) Transportation Problem: (Vogel’s Approximation Method)
The North-West Corner method and the Least Cost Cell method has been discussed
in the previous session. In this session, the Vogel’s Approximation method will be discussed.
Solution:
 For each row find the least value and then the second least value and take the
absolute difference of these two least values and write it in the corresponding row
difference as shown in the image below. In row O1, 1 is the least value and 3 is the
second least value and their absolute difference is 2. Similarly, for row O2 and O3,
the absolute differences are 3 and 1 respectively.
 For each column find the least value and then the second least value and take the
absolute difference of these two least values then write it in the corresponding
column difference as shown in the figure. In column D1, 2 is the least value and 3 is
the second least value and their absolute difference is 1. Similarly, for column D2,
D3and D3, the absolute differences are 2, 2 and 2 respectively.
 These value of row difference and column difference are also called as penalty. Now
select the maximum penalty. The maximum penalty is 3 i.e. row O2. Now find the
cell with the least cost in row O2 and allocate the minimum among the supply of the
respective row and the demand of the respective column. Demand is smaller than
the supply so allocates the column’s demand i.e. 250 to the cell. Then cancel the
column D1.
 From the remaining cells, find out the row difference and column difference.
 Again, select the maximum penalty which is 3 corresponding to row O1. The least-
cost cell in row O1 is (O1, D2) with cost 1. Allocate the minimum among supply and
demand from the respective row and column to the cell. Cancel the row or column
with zero value.
 Now find the row difference and column difference from the remaining cells.
 Now select the maximum penalty which is 7 corresponding to column D4. The least
cost cell in column D4 is (O3, D4) with cost 2. The demand is smaller than the supply
for cell (O3, D4). Allocate 200 to the cell and cancel the column.
 Find the row difference and the column difference from the remaining cells.
 Now the maximum penalty is 3 corresponding to the column D2. The cell with the
least value in D2 is (O3, D2). Allocate the minimum of supply and demand and cancel
the column
 Now there is only one column so select the cell with the least cost and allocate the
value
 Now
there is
only one cell so allocate the remaining demand or supply to the cell
 No balance remains. So, multiply the allocated value of the cells with their
corresponding cell cost and add all to get the final cost i.e. (300 * 1) + (250 * 2) + (50
* 3) + (250 * 3) + (200 * 2) + (150 * 5) = 2850
d) Transportation Problem: Unbalanced problem
In this session, the method to solve the unbalanced transportation problem will be
discussed. Below transportation problem is an unbalanced transportation problem.
The problem is unbalanced because the sum of all the supplies i.e. O1, O2, O3 and
Solution:
In this type of problem, the concept of a dummy row or a dummy column will be
used. As in this case, since the supply is more than the demand so a dummy demand column
will be added and a demand of (total supply – total demand) will be given to that column i.e.
117 – 95 = 22 as shown in the image below. If demand were more than the supply then a
dummy supply row would have been added.
Now that the problem has been updated to a balanced transportation problem, it
can be solved using any one of the following methods to solve a balanced transportation
problem as discussed in the earlier posts:
1. Northwest Corner Method
2. Least Cost Cell Method
3. Vogel’s Approximation Method
Optimal solution: MODI Method – UV Method
There are two phases to solve the transportation problem. In the first phase, the
initial basic feasible solution has to be found and the second phase involves optimization of
the initial basic feasible solution that was obtained in the first phase.
There are three methods for finding an initial basic feasible solution:
1. Northwest Corner Method
2. Least Cost Cell Method
3. Vogel’s Approximation Method Will discuss how to optimize the initial basic
feasible solution through an explained example.
Consider the below transportation problem.
Solution:
Step 1: Check whether the problem is balanced or not. If the total sum of all the supply from
sources O1, O2, and O3 is equal to the total sum of all the demands for destinations D1, D2,
D3 and D4 then the transportation problem is a balanced transportation problem.
Note: If the
problem is not unbalanced then the concept of a dummy row or a dummy column to
transform the unbalanced problem to balanced can be followed as discussed. Step 2:
Finding the initial basic feasible solution. Any of the three aforementioned methods can be
used to find the initial basic feasible solution. Here, Northwest Corner Method will be used.
And according to the Northwest Corner Method this is the final initial basic feasible solution:
Now, the total cost of transportation will be (200 * 3) + (50 * 1) + (250 * 6) + (100 *
5) + (250 * 3) + (150 * 2) = 3700.
Step 3: U-V method to optimize the initial basic feasible solution
The following is the initial basic feasible solution:
– For U-V method the

values ui and vj have to be found for the rows and the columns respectively. As there are
three rows so three ui values have to be found i.e. u1 for the first row, u2 for the second
row and u3 for the third row.
Similarly, for four columns four vj values have to be found i.e. v1, v2, v3 and v4.
Check the image below:
There is a separate formula to find ui and vj, ui + vj = Cij where Cij is the cost value
only for the allocated cell. Before applying the above formula, we need to check whether m
+ n – 1 is equal to the total number of allocated cells or not where m is the total number of
rows and n is the total number of columns.
In this case m = 3, n = 4 and total number of allocated cells is 6 so m + n – 1 = 6. The
case when m + n – 1 is not equal to the total number of allocated cells will be discussed in
the later posts. Now to find the value for u and v we assign any of the three u or any of the
four v as 0. Let we assign u1 = 0 in this case. Then using the above formula, we will get v1 = 3
as u1 + v1 = 3 (i.e. C11) and v2 = 1 as u1 + v2 = 1 (i.e. C12). Similarly, we have got the value for
v2 = 3 so we get the value for u2 = 5 which implies v3 = 0. From the value of v3 = 0 we get
u3 = 3 which implies v4 = -1.
Now, compute penalties using the formula Pij = ui + vj – Cij only for unallocated cells.
We have two unallocated cells in the first row, two in the second row and two in the third
row. Let’s compute this one by one.
1. For C13, P13 = 0 + 0 – 7 = -7 (here C13 = 7, u1 = 0 and v3 = 0)
2. For C14, P14 = 0 + (-1) -4 = -5
3. For C21, P21 = 5 + 3 – 2 = 6
4. For C24, P24 = 5 + (-1) – 9 = -5
5. For C31, P31 = 3 + 3 – 8 = -2
6. For C32, P32 = 3 + 1 – 3 = 1
The Rule: If we get all the penalties value as zero or negative values that mean the
optimality is reached and this answer is the final answer. But if we get any positive value
means we need to proceed with the sum in the next step. Now find the maximum positive
penalty. Here the maximum value is 6 which corresponds to C21 cell. Now this cell is new
basic cell. This cell will also be included in the solution.
The rule for
drawing closed-path or loop. Starting from the new basic cell draw a closed-path in such a
way that the right-angle turn is done only at the allocated cell or at the new basic cell.
Assign alternate plus-minus sign to all the cells with right angle turn (or the corner)
in the loop with plus sign assigned at the new basic cell.
Consider the cells with a negative sign. Compare the allocated value (i.e. 200 and 250
in this case) and select the minimum (i.e. select 200 in this case). Now subtract 200 from the
cells with a minus sign and add 200 to the cells with a plus sign. And draw a new iteration.
The work of the loop is over and the new solution looks as shown below.
Check the total number of allocated cells is equal to (m + n – 1). Again, find u values and v
values using the formula ui + vj = Cij where Cij is the cost value only for allocated cell. Assign u1 = 0
then we get v2 = 1. Similarly, we will get following values for ui and vj.
Find the penalties for all the unallocated cells using the formula Pij = ui + vj – Cij.
1. For C11, P11 = 0 + (-3) – 3 = -6
2. For C13, P13 = 0 + 0 – 7 = -7
3. For C14, P14 = 0 + (-1) – 4 = -5
4. For C24, P24 = 5 + (-1) – 9 = -5
5. For C31, P31 = 0 + (-3) – 8 = -11
6. For C32, P32 = 3 + 1 – 3 = 1
There is one positive value i.e. 1 for C32. Now this cell becomes new basic cell.
Now draw a loop starting from the new basic cell. Assign alternate plus and minus sign with
new basic cell assigned as a plus sign.
Check if the total number of allocated cells is equal to (m + n – 1). Find u and v values as
above.
Now again find the penalties for the unallocated cells as above.
1. For P11 = 0 + (-2) – 3 = -5
2. For P13 = 0 + 1 – 7 = -6
3. For P14= 0 + 0 – 4 = -4
4. For P22= 4 + 1 – 6 = -1
5. For P24= 4 + 0 – 9 = -5
6. For P31= 2 + (-2) – 8 = -8 All the penalty values are negative values.
So, the optimality is reached. Now, find the total cost i.e. (250 * 1) + (200 * 2) + (150
* 5) + (50 * 3) + (200 * 3) + (150 * 2) = 2450
Transportation Problem: Degeneracy in Transportation Problem
This session will discuss degeneracy in transportation problem through an explained example.
Solution:
This problem is balanced transportation problem as total supply is equal to total demand.
Initial basic feasible solution:
Least Cost Cell Method will be used here to find the initial basic feasible solution.
One can also use Northwest Corner Method or Vogel’s Approximation Method to find the
initial basic feasible solution. Using Least Cost Cell Method, we get the following solution.
Optimization of the solution using U-V Method:

Check whether m + n – 1 = total number of allocated cells. In this case m + n – 1 = 4 + 5 – 1 =
8 whereas total number of allocated cells are 7, hence this is the case of degeneracy in
transportation problem. So, in this case we convert the necessary number (in this case it is
m + n – 1 – total number of allocated cells i.e. 8 – 7 = 1) of unallocated cells into allocated
cells to satisfy the above condition.
Steps to convert unallocated cells into allocated cells:
 Start from the least value of the unallocated cell.
 Check the loop formation one by one.
 There should be no closed-loop formation.
 Select that loop as a new allocated cell and assign a value ‘e’.
The closed loop can be in any form but all the turning point should be only at
allocated cell or at the cell from the loop is started.
There are 13 unallocated cells. Select the least value (i.e. 5 in this case) from
unallocated cells. There are two 5s here so you can select randomly anyone. Let’s select the
cell with star marked.
Check if there is any closed-loop formation starting from this cell. If a closed-loop is
drawn from this cell following the condition for closed-loop then it can be observed that this
cell cannot be reached to complete the closed-loop. So, this cell will be selected and
assigned a random value ‘e’.
Note: If the closed loop would have been formed from that cell then we would try another
cell with least value and do the same procedure and check whether closed loop is possible
or not. Now total number of allocated cells becomes 8 and m + n – 1 = 4 + 5 – 1 = 8. Now
this solution can be optimized using U-V method. We get the below solution after
performing optimization using U-V method.
The presence of two ‘e’ in the final solution means after doing some iterations during
optimization, the condition for degeneracy will be met once again. While finding the total
cost, just leave the ‘e’ and multiply the allocated value with its cell’s cost value and add all of
them. So, the transportation cost is (35 * 3) + (20 * 5) + (10 * 2) + (10 * 4) + (20 * 5) + (5 *
13) + (25 * 8) = 630.
ASSIGNMENT PROBLEMS
INTRODUCTION
In Block 1 of this course, we have discussed the basic concepts elated to Linear
Programming Problems and the Simplex method for solving them. The Transportation
Problem was also discussed in Block 1. In this unit, we explain the Assignment problem and
discuss various methods for solving it. The assignment problem deals with allocating various
resources (items) to various activities (receivers) on a one to one basis, i.e., the number of
operations is to be assigned to an equal number of operators where each operator performs
only one operation. For example, suppose an accounts officer has 4 subordinates and 4
tasks. The subordinates differ in efficiency and take different time to perform each task. If
one task is to be assigned to one person in such a way that the total person hours are
minimized, the problem is called an assignment problem. Though the assignment problem is
a special case of transportation problem, it is not solved using the methods described in
previous lessons. We use another method called the Hungarian method for solving an
assignment problem. It is shorter and easier compared to any method of finding the optimal
solution of a transportation problem. In this unit, we discuss various types of assignment
problems, including travelling salesman problem and apply the Hungarian method for
solving these problems. In the next unit, we shall discuss the fundamental structure and
operating characteristics of a queueing system and explain a single server M/M/1 queueing
model with Poisson input and exponential service time.
An assignment problem may be considered as a special type of transportation
problem in which the number of sources and destinations are equal. The capacity of each
source as well as the requirement of each destination is taken as 1. In the case of an
assignment problem, the given matrix must necessarily be a square matrix which is not the
condition for a transportation problem.
Suppose there are n persons and n jobs and the assignment of jobs has to be done
on a one-to-one basis. This assignment problem can be stated in the form of an n × n matrix
of real numbers (known as the cost matrix) as given in the following table:
xi1  xi2 ... xin 1, i  1, 2,..., n (one job is done by theith person)
x1j  x2j...  xnj 1, j 1,2,….. n (only one person is assigned the jth job)
The constant cij in the above problem represents time. It may be cost or some other
parameter which is to be minimized in the assignment problem under consideration.
Note that an assignment problem is a special type of transportation problem and may be
solved as one. However, we use another method known as the Hungarian method for
solving it. This method is shorter and easier compared to any other method of finding the
optimal solution of a transportation problem. Let us explain the Hungarian method of
finding the optimal solution of an assignment problem.
Hungarian Method of Solving an Assignment Problem
The steps for obtaining an optimal solution of an assignment problem are as follows:
1. Check whether the given matrix is square. If not, make it square by adding a suitable
number of dummy rows (or columns) with 0 cost/time elements.
2. Locate the smallest cost element in each row of the cost matrix. Subtract the smallest
element of each row from every element of that row.
3. In the resulting cost matrix, locate the smallest element in each column and subtract the
smallest element of each column from every element of that column.
4. In the resulting matrix, search for an optimum assignment as follows:
i) Examine the rows successively until a row with exactly one zero is found. Draw a
rectangle around this zero (as 0) and cross out all other zeroes in the corresponding column.
Proceed in this manner until all the rows have been examined. If there is more than one
zero in any row, do not touch that row; pass on to the next row.
ii) Repeat step (i) above for the columns of the resulting cost matrix.
iii) If a row or column of the reduced matrix contains more than one zeroes,
arbitrarily choose a row or column having the minimum number of zeroes. Arbitrarily select
any zero in the row or column so chosen. Draw a rectangle around it and cross out all the
zeroes in the corresponding row and column. Repeat steps (i), (ii) and (iii) until all the zeroes
have either been assigned (by drawing a rectangle around them) or crossed.
iv) If each row and each column of the resulting matrix has one and only one
assigned 0, the optimum assignment is made in the cells corresponding to 0 . The optimum
solution of the problem is attained and you can stop here.
Otherwise, go to the next step.
5. Draw the minimum number of horizontal and/or vertical lines through all the zeroes as
follows:
i) Tick mark (  ) the rows in which assignment has not been made.
ii) Tick mark (  ) columns, which have zeroes in the marked rows.
iii) Tick mark (  ) rows (not already marked) which have assignments in marked
columns. Then tick mark (  ) columns, which have zeroes in newly marked rows, if any. Tick
mark (  ) rows (not already marked), which have assignments in these newly marked
columns.
iv) Draw straight lines through all unmarked rows and marked columns.
6. Revise the cost matrix as follows:
i) Find the smallest element not covered by any of the lines.
ii) Subtract this from all the uncovered elements and add it to the elements at the
intersection of the two lines.
iii) Other elements covered by the lines remain unchanged.
7. Repeat the procedure until an optimum solution is attained.
We now illustrate the procedure with the help of an example.
Example 1: A computer center has four expert programmers and needs to develop four
application programs. The head of the computer center, estimates the computer time (in
minutes) required by the respective experts to develop the application programs as follows:
Solution: Let us subtract the minimum element of each row from every element of that row.
Note that the minimum element in the first row is 80. So, 80 is to be subtracted from every
element of the first row, i.e., from 120, 100, 80 and 90, respectively. As a result, the
elements of the first row of the resulting matrix would be 40, 20, 0, 10, respectively.
Similarly, we obtain the elements of the other rows of the resulting matrix. Thus, the
resulting matrix is:
Let us now subtract the minimum element of each column from every element of
that column in the resulting matrix. The minimum element in the first column is 10. So, 10 is
to be subtracted from every element of the first column, i.e., from 40, 10, 10, and 10,
respectively. As a result, the elements of the first column of the resulting matrix are 30, 0, 0,
0, respectively. Similarly, we obtain the elements of the other columns of the resulting
matrix. Thus, the resulting matrix is:
In the second, third and fourth row, there is no single zero. Hence, we move column-
wise. In the second column, we have a single zero. Hence, we draw a rectangle around it
and cross all other zeroes in the corresponding row. We get
In the matrix above, there is no row or column, which has a single zero. Therefore,
we first move row-wise to locate the row having more than one zero. The second row has
two zeroes. So, we draw a rectangle arbitrarily around one of these zeroes and cross the
other one. Let us draw a rectangle around the zero in the cell (2, A) and cross the zero in the
cell (2, D). We cross out the other zeroes in the first column. Note that we could just as well
have selected the zero in the cell (2, D), drawn a rectangle around it and crossed all other
zeroes. This would have led to an alternative solution.
In this way, we are left with only one zero in every row and column around which a
rectangle has been drawn. This means that we have assigned only one operation to one
operator. Thus, we get the optimum solution as follows:
This means that programmer 1 is assigned programme C, programmer 2 is assigned

programme A, and so on. The minimum time taken in developing the programmes is
= 80 + 80 + 100 + 90 = 350 min.
Example 2: A company is producing a single product and selling it through five agencies
situated in different cities. All of a sudden, there is a demand for the product in five more
cities that do not have any agency of the company.
The company is faced with the problem of deciding on how to assign the existing
agencies to dispatch the product to the additional cities in such a way that the travelling
distance is minimized. The distances (in km) between the surplus and deficit cities are given
in the following distance matrix.
Since the number of assignments is less than the number of rows (or columns), we proceed
from Step 5 onwards of the Hungarian method as follows:
i) we tick mark (  ) the rows in which the assignment has not been made.
These are the 3rd and 5th rows.
ii) ii) we tick mark (  ) the columns which have zeroes in the marked rows. This
is the 2nd column.
iii) iii) we tick mark (  ) the rows which have assignments in marked columns.
This is the 1st row.
iv) iv) again, we tick mark (  ) the column(s) which have zeroes in the newly
marked row. This is the 2nd column, which has already been marked. There is
no other such column. So, we have
We draw straight lines through unmarked rows and marked columns as follows:
A  V, B  III, C  II, D  I, E  IV
Thus, the minimum distance is 200 + 130 + 110 + 50 + 80 = 570 km. You should pause here
and try to solve the following assignment problem to check your understanding.
POST-TEST:
Solve the following:
1. Wheat is harvested in the Midwest and stored in grain elevators in three different
cities – Kansas City, Omaha, and Des Moines. These grain elevators supply three
flour mills, located in Chicago, St. Louis, and Cincinnati. Grain is shipped to the mills
in railroad cars, each car capable of holding one ton of wheat. Each grain elevator is
able to supply the following number of tons (i.e., railroad cars) of wheat to the mills
on monthly basis.
3.
4.
LESSON 5: DECISION THEORY
PRE-TEST:
1. What is a Decision Theory?
2. How to construct a Decision Tree?
1. Describe the barriers to individual decision-making and common styles of decision-
making.
2. Explain the concept of “rational decision making” and contrast it with prospect
theory, bounded rationality, heuristics, and robust decisions.
3. Explain evidence-based decision making and its tools.
4. Describe the components and use of a decision tree.
5. Explain common techniques used to manage group decision making.
What is Decision Theory?

Decision theory is the study of a person or agents’ choices. The theory helps us understand
the logic behind the choices professionals, consumers, or even voters make.

The choices come with consequences and are usually discussed in two separate but
distinct branches. The branches consist of Normative Decision Theory and Optimal Decision
Theory.
When analyzing decision theory, the analysis often consists of what makes an
optimal decision, who that optimal decision-maker may be, and how they can come to that
decision. Discussing how people “ought” to make decisions in certain scenarios is part of this
study as well.
We will discuss how the two fields differ and how decision theory plays a part in our
everyday lives in many different facets. The field of decision theory is often studied in
business schools, with appearances in the field of economics and statistics. It can also be
found in the scientific realm, with biologists and political scientists studying it as well.
Normative and Optimal Decision Theory Analyses
Decision theory is divided into two different subcategories. The table below
discusses and outlines some of the critical differences between the two, and the respective
areas they focus on:
Normative Decision Theory Optimal Decision Theory
Analyzes the outcomes of decisions The investigation and analysis of why individuals and
agents of choice make the decisions that they do
Determines the optimal decision Looks at the assumptions made by individuals making
based on outcomes the decisions and the assumptions they make when
deciding
The theory centers around the ideal

decision-maker for a specific situation
"What should the person making the

decision do to make this decision?"
Decision theory
Decision theory is the study of how people should combine their values and
understanding of the world to select actions or strategies. Decision theory is normative: it
provides us with a means of assessing decisions, and guidance on how to make the best
decisions. The goal is not necessarily to show how real-life agents actually make decisions,
which is explored in behavioral economics and the psychology of decision making.
Suppose that a person has more than one action available to them, that they
are uncertain about the state of the world, and that they know that different outcomes will
result from their action depending on what the true state of the world is. Even once they
have assigned values to each of these outcomes, they will need a function to process those
value-assigned outcomes to the actions available to them: one that can rank the actions by
how good they are. For example, if they think that they should maximize expected value,
then the function will rank all of the actions available to them by the amount of value that
each of them is expected to produce.
Decision theory, in statistics, a set of quantitative methods for reaching optimal
decisions. A solvable decision problem must be capable of being tightly formulated in terms
of initial conditions and choices or courses of action, with their consequences. In general,
such consequences are not known with certainty but are expressed as a set of probabilistic
outcomes. Each outcome is assigned a “utility” value based on the preferences of the
decision maker. An optimal decision, following the logic of the theory, is one that maximizes
the expected utility. Thus, the ideal of decision theory is to make choices rational by
reducing them to a kind of routine calculation.
Attempting to Predict Consumer Behavior
Understanding how decision theory works and its implications for consumer
behavior is an excellent tool for marketers to utilize. Understanding decision theory can help
when making predictive models for which item a consumer may choose or even which type
of transportation they may utilize. It can help in understanding alternatives an individual
may choose and why they may choose them.
Understanding consumer behavior can be pivotal in determining whether a product
will be a success. After all, creating a quality product is not necessarily indicative of whether
or not people will buy it and whether a company will be able to recover the investment
made in R&D and marketing they put forth trying to sell it.
Why are decision making styles important?
Whether you are leading a team, a corporation or evaluating your own tasks, it's
important to understand how you make decisions because knowing how you and your
coworkers make a decision can lead to a more productive and cooperative work
environment. Practicing good decision-making can improve your leadership qualities.
Getting to know the four decision-making styles will help you understand your own process
and the way others make choices in the workplace. When you identify your own decisive
style, you can learn how to manage the outcomes of a situation when you need to provide a
solution.
Four decision-making styles (with examples)
Each decision-making style is characterized by either a task or social focus and a high
or low tolerance for ambiguity. Styles with a high tolerance for ambiguity can work with
unknown variables as they come to a conclusion. Those with a low tolerance for ambiguity
want as much clarity as possible in all the circumstances and information that lead to their
decisions.
Decision-making styles also vary in a social or task-driven focus. Social-driven
decisions consider the behavior of others involved in the outcome. Those who are task-
driven make decisions based on how to best achieve a goal.
Here are the four decision-making styles with examples of how they might be used in the
workplace:
1. Directive
The directive decision-making style uses quick, decisive thinking to come to a
solution. A directive decision-maker has a low tolerance for unclear or ambiguous ideas.
They are focused on the task and will use their own knowledge and judgment to come to a
conclusion with selective input from other individuals.
Directive decision-makers excel at verbal communication. They are rational and
logical in their decision making. When the team or organization needs a fast decision, a
directive-style decision-maker can effectively make a choice. Their style is valuable for
making short-term decisions.
Example: Company stockholders have voted to expand their 401(k) option to all current
employees and new employees after they complete a 90-day trial period. The CEO must now
decide if the company will provide matching funds for employees who give to their 401(k)
fund. She thinks about how this might help to attract top talent for their team.
The CE
O looks at the budget projections she has just prepared and thinks about how funds that are
allocated for another project could be used to match employee contributions. She decides
that employees who contribute to their funds will be matched 4% by the company.
2. Analytical
Analytical decision-makers carefully analyze data to come up with a solution. They
are careful and adaptable thinkers. They will invest time to glean information to form a
conclusion. These decision-makers are task-oriented, but have a high tolerance for
ambiguity.
Analytical decision-makers take time to compile data and evidence before they come
to a conclusion. When they do make a decision, they have looked at all the details and
formed what they believe is the best possible solution.
Example: The marketing team of a sports broadcasting company is tasked to identify how
they can reach a wider audience with their current ad campaign. The marketing manager
asks each team leader to submit a report from their portion of the campaign including the
numbers of each audience demographic. They read each report then meet with the team
leads. After the meeting, the marketing manager decides to purchase more add space on
social media websites for the next 30 days.
3. Conceptual
Those who make decisions with a conceptual style are big picture thinkers who are
willing to take risks. They evaluate different options and possibilities with a high tolerance to
ambiguity. They are social-oriented and take time to consider big ideas and creative
solutions.
Conceptual decision-makers look forward to what could happen if the decision is
made. Their conclusions come from visualizing different opportunities and outcomes for the
future. They are strong in making long-term decisions.
Example: Joe's startup retail company is performing well during their first year. He thinks
about how the company can open stores nationwide in the next five years. When a new
shopping development begins construction in a nearby big city, he decides to open a new
store branch at the site. Although it is a risk to open this new store, Joe is confident his team
will be successful, and this will help launch their brand nationally.
4. Behavioral
A behavioral style of decision-making focuses on relationships more than the task. It
evaluates the feelings of others as part of their decision-making process. Behavior decision-
makers have a low tolerance for ambiguity and a social focus as they evaluate solutions.
These decision-makers rely on information from others to guide what they choose.
They are persuasive communicators who value decisions based on a team consensus. Their
decisions are often based on how the choice will impact relationships.
Example: As HR manager, Kate has been asked to decide which week employees should get
as a bonus paid vacation days before the end of the year. She sends out an email survey to
see how employees feel about three possible dates. After she reads the survey responses,
she asks her coworkers for input over a lunch break. Later in the afternoon, she walks
through the office chatting with several more employees.
When she decides which week the majority of coworkers want, she talks to several
employees who will not get the choice they hoped for, making sure they're feeling alright
about the decision. At the end of the day, she notifies the management and the employees
which week will be a bonus paid vacation time.
Types of decisions
1. Choice under uncertainty
The area of choice under uncertainty represents the heart of decision theory. Known
from the 17th century, the idea of expected value is that, when faced with a number of
actions, each of which could give rise to more than one possible outcome with different
probabilities, the rational procedure is to identify all possible outcomes, determine their
values (positive or negative) and the probabilities that will result from each course of action,
and multiply the two to give an "expected value", or the average expectation for an
outcome; the action to be chosen should be the one that gives rise to the highest total
expected value.
2. Intertemporal choice
Intertemporal choice is concerned with the kind of choice where different actions
lead to outcomes that are realized at different stages over time It is also described as cost-
benefit decision making since it involves the choices between rewards that vary according to
magnitude and time of arrival. If someone received a windfall of several thousand dollars,
they could spend it on an expensive holiday, giving them immediate pleasure, or they could
invest it in a pension scheme, giving them an income at some time in the future. What is the
optimal thing to do? The answer depends partly on factors such as the expected rates of
interest and inflation, the person's life expectancy, and their confidence in the pensions
industry. However even with all those factors taken into account, human behavior again
deviates greatly from the predictions of prescriptive decision theory, leading to alternative
models in which, for example, objective interest rates are replaced by subjective discount
rates.
3. Interaction of decision makers
Some decisions are difficult because of the need to take into account how other
people in the situation will respond to the decision that is taken. The analysis of such social
decisions is more often treated under the label of game theory, rather than decision theory,
though it involves the same mathematical methods. From the standpoint of game theory,
most of the problems treated in decision theory are one-player games (or the one player is
viewed as playing against an impersonal background situation). In the emerging field
of socio-cognitive engineering, the research is especially focused on the different types of
distributed decision-making in human organizations, in normal and
abnormal/emergency/crisis situations.
4. Complex decisions
Other areas of decision theory are concerned with decisions that are difficult simply
because of their complexity, or the complexity of the organization that has to make them.
Individuals making decisions are limited in resources (i.e. time and intelligence) and are
therefore boundedly rational; the issue is thus, more than the deviation between real and
optimal behavior, the difficulty of determining the optimal behavior in the first place. One
example is the model of economic growth and resource usage developed by the Club of
Rome to help politicians make real-life decisions in complex situations. Decisions are also
affected by whether options are framed together or separately; this is known as
the distinction bias.
Understanding Decision Theory
Decision theory brings together psychology, statistics, philosophy, and mathematics
to analyze the decision-making process. Decision theory is closely related to game
theory and is studied within the context of understanding the activities and decisions
underpinning activities such as auctions, evolution, and marketing.
There are three main areas of decision theory. Each studies a different type of decision
making.
1. Descriptive decision theory: describes the typical manner in which agents make
decisions---some of it rational, some of it intuitive, some of it quite irrational.
2. Prescriptive decision theory: tries to provide guidelines for agents to make the best
possible decisions given an uncertain decision-making framework.
3. Normative decision theory: provides guidance for making decisions given a set of
values.
Decision theory framework generally identifies three types of decision classes:
1. Decisions under certainty: an abundance of information leads to an obvious decision.
2. Decisions under uncertainty: analysis of known and unknown variables lead to the
best probabilistic decision.
3. Decisions under conflict: a reactive approach that involves anticipating potential
consequences to the decision, prior to making a decision.
Decision Under Uncertainty: Prisoner's Dilemma
A common example of decision theory stems from the prisoner's dilemma in which
two individuals are faced with an uncertain decision where the outcome is not only based
on their personal decision, but also on that of the other individual. Since both parties do not
know what actions the other person will take, this results in an uncertain decision
framework. While mathematics and statistical models determine what the optimal decision
should be, psychology and philosophy introduce factors of human behaviors to suggest the
most likely outcome.
The Uncertainty Variable in Decision Theory
Different uncertainty variables are a part of the decision-making theory. Former
university professor Leonard Jimmie Savage, the author of “The Foundations of Statistics,”
outlined the different conditions of uncertainty that exist in modern-day decision-making
theory.
Three different types of uncertainty can be found in decision-making theory – States,
Consequences, and Actions.
a. States encompass facts that exist in the universe that can affect a decision.
b. Consequences are the features of a decision made that influence a decision-maker
on a micro-level, i.e., whether an individual feels rested.
c. Actions are the link/bridge between both states and consequences. Savage describes
them as links between states and consequences.
Using a Decision Tree
Decision trees are useful tools, particularly for situations where financial data and
probability of outcomes are relatively reliable. They are used to compare the costs and likely
values of decision pathways that a business might take. They often include decision
alternatives that lead to multiple possible outcomes, with the likelihood of each outcome
being measured numerically.
What are decision trees used for?
Despite their drawbacks, decision trees are still a powerful and popular tool. They’re
commonly used by data analysts to carry out predictive analysis (e.g. to develop operations
strategies in businesses). They’re also a popular tool for machine learning and artificial
intelligence, where they’re used as training algorithms for supervised learning (i.e.
categorizing data based on different tests, such as ‘yes’ or ‘no’ classifiers.)
Broadly, decision trees are used in a wide range of industries, to solve many types of
problems. Because of their flexibility, they’re used in sectors from technology and health to
financial planning. Examples include:
 A technology business evaluating expansion opportunities based on analysis of past
sales data.
 A toy company deciding where to target its limited advertising budget, based on
what demographic data suggests customers are likely to buy.
 Banks and mortgage providers using historical data to predict how likely it is that a
borrower will default on their payments.
 Emergency room triage might use decision trees to prioritize patient care (based on
factors such as age, gender, symptoms, etc.)
 Automated telephone systems guiding you to the outcome you need, e.g. ‘For option
A, press 1. For option B, press 2’, and so on.
As you can see, there many uses for decision trees!
Steps in Decision Tree Analysis
1. The first
step is understanding and specifying the problem area for which decision making is
required.
2. The second step is interpreting and chalking out all possible solutions to the
particular issue as well as their consequences.
3. The third step is presenting the variables on a decision tree along with its respective
probability values.
4. The fourth step is finding out the outcomes of all the variables and specifying it in the
decision tree.
5. The last step is highly crucial and backs the overall analysis of this process. It involves
calculating the EMV values for all the chance nodes or options, to figure out the
solution which provides the highest expected value.
Advantages of decision trees
 Good for interpreting data in a highly visual way.
 Good for handling a combination of numerical and non-numerical data.
 Easy to define rules, e.g. ‘yes, no, if, then, else…’
 Requires minimal preparation or data cleaning before use.
 Great way to choose between best, worst, and likely case scenarios.
 Can be easily combined with other decision-making techniques.
Disadvantages of decision trees
 Overfitting (where a model interprets meaning from irrelevant data) can become a
problem if a decision tree’s design is too complex.
 They are not well-suited to continuous variables (i.e. variables which can have more
than one value, or a spectrum of values).
 In predictive analysis, calculations can quickly grow cumbersome, especially when a
decision path includes many chance variables.
 When using an imbalanced dataset (i.e. where one class of data dominates over
another) it is easy for outcomes to be biased in favor of the dominant class.
 Generally, decision trees provide lower prediction accuracy compared to other
predictive algorithms.
What are the different parts of a decision tree?
Decision trees can deal with complex data, which is part of what makes them useful.
However, this doesn’t mean that they are difficult to understand. At their core, all decision
trees ultimately consist of just three key parts, or ‘nodes’:
 Decision nodes: Representing a decision (typically shown with a square)
 Chance nodes: Representing probability or uncertainty (typically denoted by a circle)
 End nodes: Representing an outcome (typically shown with a triangle)
Connecting these different nodes are what we call branches. Nodes and branches can be
used over and over again in any number of combinations to create trees of various
complexity. Let’s see how these parts look before we add any data.
Luckily, a lot of decision tree terminology follows the tree analogy, which makes it much
easier to remember! Some other terms you might come across will include:
Root nodes
In the diagram above, the blue decision node is what we call a ‘root node.’ This is
always the first node in the path. It is the node from which all other decision, chance, and
end nodes eventually branch.
Leaf nodes
In the diagram above, the lilac end nodes are what we call ‘leaf nodes.’ These show
the end of a decision path (or outcome). You can always identify a leaf node because it
doesn’t split, or branch any further. Just like a real leaf!
Internal nodes
Between the root node and the leaf nodes, we can have any number of internal
nodes. These can include decisions and chance nodes (for simplicity, this diagram only uses
chance nodes). It’s easy to identify an internal node—each one has branches of its own
while also connecting to a previous node.
Splitting
Branching or ‘splitting’ is what we call it when any node divides into two or more
sub-nodes. These sub-nodes can be another internal node, or they can lead to an outcome
(a leaf/ end node.)
Pruning
Sometimes decision trees can grow quite complex. In these cases, they can end up
giving too much weight to irrelevant data. To avoid this problem, we can remove certain
nodes using a process known as ‘pruning’. Pruning is exactly what it sounds like—if the tree
grows branches, we don’t need, we simply cut them off. Easy!
An example of a simple decision tree
Now that we’ve covered the basics, let’s see how a decision tree might look. We’ll
keep it really simple. Let’s say that we’re trying to classify what options are available to us if
we are hungry. We might show this as follows:
In
this diagram, our different options are laid out in a clear, visual way. Decision nodes are
navy blue, chance nodes are light blue, and end nodes are purple. It is easy for anybody to
understand and to see the possible outcomes.
However, let’s not forget: our aim was to classify what to do in the event of being
hungry. By including options for what to do in the event of not being hungry, we’ve
overcomplicated our decision tree. Cluttering a tree in this way is a common problem,
especially when dealing with large amounts of data. It often results in the algorithm
extracting meaning from irrelevant information. This is known as overfitting. One option to
fix overfitting is simply to prune the tree:
As you can see, the focus of our decision tree is now much clearer. By removing the
irrelevant information (i.e. what to do if we’re not hungry) our outcomes are focused on the
goal we’re aiming for. This is one example of a pitfall that decision trees can fall into, and
how to get around it. However, there are several pros and cons for decision trees. Let’s
touch on these next.
Probability and Decision Making - Value and Utility
The future is uncertain, but we have to make decisions every day that have an effect
on our prospects, financial and otherwise. Faced with uncertainty, we do not merely throw
up our hands and guess randomly about what to do; instead, we assess the potential risks
and benefits of a variety of options, and choose to act in a way that maximizes the
probability of a beneficial outcome. Things won’t always turn out for the best, but we have
to try to increase the chances that they will. To do so, we use our knowledge—or at least
our best estimates—of the probabilities of future events to guide our decisions.
The process of decision-making in the face of uncertainty is most clearly illustrated
with examples involving financial decisions. When we make a financial investment, or—
what’s effectively though not legally the same thing—a wager, we’re putting money at risk
with the hope that it will pay off in a larger sum of money in the future. We need a way of
deciding whether such bets are good ones or not. Of course, we can evaluate our financial
decisions in hindsight, and deem the winning bets good choices and the losing one’s bad
choices, but that’s not a fair standard. The question is, knowing what we knew at the time
we made our decision, did we make the choice that maximized the probability of a
profitable outcome, even if the profit was not guaranteed?
To evaluate the soundness of a wager or investment, then, we need to look not at its
worth after the fact—its final value, we might say—but rather at the value we can
reasonably expect it to have in the future, based on what we know at the time the decision
is made. We’ll call this the expected value (sometimes called expectation value).
What is Expected Value?
Expected value (also known as EV, expectation, average, or mean value) is a long-run
average value of random variables. It also indicates the probability-weighted average of all
possible values.
Expected value is a commonly used financial concept. In finance, it indicates the
anticipated value of an investment in the future. By determining the probabilities of possible
scenarios, one can determine the EV of the scenarios. The concept is frequently used with
multivariate models and scenario analysis. It is directly related to the concept of expected
return.
Formula for Expected Value
The first variation of the expected value formula is the EV of one event repeated several
times (think about tossing a coin). In such a case, the EV can be found using the following
formula:

Where:
EV – the expected value
P(X) – the probability of the event
n – the number of the repetitions of the event
However, in finance, many problems related to the expected value involve multiple events.
In such a scenario, the EV is the probability-weighted average of all possible events.
Therefore, the general formula to find the EV for multiple events is:
Where:
EV – the expected value
P(XI) – the probability of the event
XI – the event
Example of Expected Value (Multiple Events)

1. You are a financial analyst in a development company. Your manager just asked you
to assess the viability of future development projects and select the most promising one.
According to estimates, Project A, upon completion, shows a probability of 0.4 to achieve a
value of $2 million and a probability of 0.6 to achieve a value of $500,000. Project B shows a
probability of 0.3 to be valued at $3 million and a probability of 0.7 to be valued at $200,000
upon completion.

In order to select the right project, you need to calculate the expected value of each project
and compare the values with each other. The EV can be calculated in the following way:
EV (Project A) = [0.4 × $2,000,000] + [0.6 × $500,000] = $1,100,000
EV (Project B) = [0.3 × $3,000,000] + [0.7 × $200,000] = $1,040,000
 The EV of Project A is greater than the EV of Project B. Therefore, your
company should select Project A.
 Note that the example above is an oversimplified one. A real-life example will
likely assess the Net Present Value (NPV) of the projects instead of their EV.
However, NPV calculations also consider the EV of different projects.
2. A normal six-sided die. Once you roll the die, it has an equal one-sixth chance of
landing on one, two, three, four, five, or six. Given this information, the calculation is
straightforward:
If you were to roll a six-sided die an infinite amount of times, you see the average value
equals 3.5.
3. ABC Ltd. is a company manufacturing skincare product. It was found that the
business is at the maturity stage, demanding some change. After rigorous research,
management came up with the following decision tree:
In the above decision tree, we can easily make out that the company can expand its
existing unit or innovate a new product, i.e., shower gel or make no changes.
Given below is the evaluation of each of these alternatives:
Expansion of Business Unit:
If the company invests in the development of its business unit, there can be two
possibilities, i.e.:
40% possibility that the market share will hike, increasing the overall profitability of the
company by ₹2500000;
60% possibility that the competitors would take over the market share and the company
may incur a loss of ₹800000.
To find out the viability of this option, let us compute its EMV (Expected Monetary Value):
New Product Line of Shower Gel:

If the organization go for new product development, there can be following two
possibilities:
50% chances are that the project would be successful and yield ₹1800000 as profit;
50% possibility of failure persists, leading to a loss of ₹800000.
To determine the profitability of this idea, let us evaluate its EMV:
Do Nothing:
If the company does not take any step, still there can be two outcomes, discussed
below:
40% chances are there that yet, the organization can attract new customers, generating a
profit of ₹1000000;
60% chances of failure are there due to the new competitors, incurring a loss of ₹400000.
Given below is the EMV in such circumstances:

Interpretation
From the above evaluation, we can easily make out that the option of a new product
line has the highest EMV. Therefore, we can say that the company can avail this opportunity
to make the highest gain by ensuring the best possible use of its resources.
4. Example
POST-TEST: Discuss item 1-4, briefly & for item 5-6, solve the ff. problem & show your
solution:
1. Discuss the branches of decision theory.
2. Why are decision making styles important?
3. Discuss the uses, advantages & disadvantages of a decision tree.
4. Discuss the three key parts of a decision tree.
5. Suppose a manager has to decide whether to accept a bid or not. If he accepts the
bid, the company may gain P2 million if it succeeds, or lose P1.5 million if it fails. The
probability that it will succeed is 30%. Find the EV if he accepts the bid.
6. Consider the problem of a student who has to decide whether to stop his studies and
work for a job paying P1,500 monthly or continue his studies, after which a job
awaits him paying P2,500 a month, provided he passes his remaining subjects. He
feels that the probability that he will pass his remaining subjects is 40%. Which
choice is better for the students?
LESSON 5: DECISION THEORY
1. Compute problems involving mathematical expectations.
2. Solve problems with perfect information.
3. To differentiate the four possible decision criterions.
4. To determine the solution set for decision under certainty, decision uncertainty and
decision risk.
5. To calculate the equivalent monetary value of decisions.
What is decision theory?

Decision theory is theory about decisions. The subject is not a very unified one. To
the contrary, there are many different ways to theorize about decisions, and therefore also
many different research traditions. This text attempts to reflect some of the diversity of the
subject. Its emphasis lies on the less (mathematically) technical aspects of decision theory.
Decision theory is the study of how people should combine their values and
understanding of the world to select actions or strategies. Decision theory is normative: it
provides us with a means of assessing decisions, and guidance on how to make the best
decisions. The goal is not necessarily to show how real-life agents actually make decisions,
which is explored in behavioral economics and the psychology of decision making.
Suppose that a person has more than one action available to them, that they
are uncertain about the state of the world, and that they know that different outcomes will
result from their action depending on what the true state of the world is. Even once they
have assigned values to each of these outcomes, they will need a function to process those
value-assigned outcomes to the actions available to them: one that can rank the actions by
how good they are. For example, if they think that they should maximize expected value,
then the function will rank all of the actions available to them by the amount of value that
each of them is expected to produce.
Decision Theory: Bayesian

Decision theory studies the logic and the mathematical properties of decision making
under uncertainty. Statistical decision theory focuses on the investigation of decision making
when uncertainty can be reduced by information acquired through experimentation. This
article reviews the Bayesian approach to statistical decision theory, as was developed from
the seminal ideas of Savage. Specifically, it considers: the principle of maximation of
expected utility and its axiomatic foundations; the basic elements of Bayesian statistical
decision theory, illustrated using standard statistical decision problems; measuring the value
of information in decision making and the concept of admissibility.
Understanding Decision Theory
Decision theory brings together psychology, statistics, philosophy, and mathematics
to analyze the decision-making process. Decision theory is closely related to game
theory and is studied within the context of understanding the activities and decisions
underpinning activities such as auctions, evolution, and marketing.
There are three main areas of decision theory. Each studies a different type of decision
making.
1. Descriptive decision theory: describes the typical manner in which agents make
decisions---some of it rational, some of it intuitive, some of it quite irrational.
2. Prescriptive decision theory: tries to provide guidelines for agents to make the best
possible decisions given an uncertain decision-making framework.
3. Normative decision theory: provides guidance for making decisions given a set of
values.
Decision theory framework generally identifies three types of decision classes:
1. Decisions under certainty: an abundance of information leads to an obvious decision
2. Decisions under uncertainty: analysis of known and unknown variables lead to the
best probabilistic decision.
3. Decisions under conflict: a reactive approach that involves anticipating potential
consequences to the decision, prior to making a decision.
Normative and descriptive theories
The distinction between normative and descriptive decision theories is, in principle,
very simple. A normative decision theory is a theory about how decisions should be made,
and a descriptive theory is a theory about how decisions are actually made. The "should" in
the foregoing sentence can be interpreted in many ways. There is, however, virtually
complete agreement among decision scientists that it refers to the prerequisites of rational
decision-making. In other words, a normative decision theory is a theory about how
decisions should be made in order to be rational. 7 This is a very limited sense of the word
"normative". Norms of rationality are by no means the only – or even the most important –
norms that one may wish to apply in decision-making. However, it is practice to regard
norms other than rationality norms as external to decision theory. Decision theory does not,
according to the received opinion, enter the scene until the ethical or political norms are
already fixed. It takes care of those normative issues that remain even after the goals have
been fixed. This remainder of normative issues consists to a large part of questions about
how to act in when there is uncertainty and lack of information. It also contains issues about
how an individual can coordinate her decisions over time and of how several individuals can
coordinate their decisions in social decision procedures. If the general wants to win the war,
the decision theorist tries to tell him how to achieve this goal. The question whether he
should at all try to win the war is not typically regarded as a decision-theoretical issue.
Similarly, decision theory provides methods for a business executive to maximize
profits and for an environmental agency to minimize toxic exposure, but the basic question
whether they should try to do these things is not treated in decision theory. Although the
scope of the "normative" is very limited in decision theory, the distinction between
normative (i.e. rationality-normative) and descriptive interpretations of decision theories is
often blurred. It is not uncommon, when you read decision-theoretical literature, to find
examples of disturbing ambiguities and even confusions between normative and descriptive
interpretations of one and the same theory. Probably, many of these ambiguities could have
been avoided. It must be conceded, however, that it is more difficult in decision science
than in many other disciplines to draw a sharp line between normative and descriptive
interpretations. This can be clearly seen from consideration of what constitutes a
falsification of a decision theory. It is fairly obvious what the criterion should be for the
falsification of a descriptive decision theory. (F1) A decision theory is falsified as a
descriptive theory if a decision problem can be found in which most human subjects
perform in contradiction to the theory. Since a normative decision theory tells us how a
rational agent should act, falsification must refer to the dictates of rationality. It is not
evident, however, how strong the conflict must be between the theory and rational
decision-making for the theory to be falsified. I propose, therefore, the following two
definitions for different strengths of that conflict. (F2) A decision theory is weakly falsified as
a normative theory if a decision problem can be found in which an agent can perform in
contradiction with the theory without being irrational. (F3) A decision theory is strictly
falsified as a normative theory if a decision problem can be found in which an agent who
performs in accordance with the theory cannot be a rational agent. Now suppose that a
certain theory T has (as is often the case) been proclaimed by its inventor to be valid both as
a normative and as a descriptive theory. Furthermore suppose (as is also often the case)
that we know from experiments that in decision problem P, most subjects do not comply
with T.
In other words, suppose that (F1) is satisfied for T. The beliefs and behaviors of
decision theoreticians are not known to be radically different from those of other human
beings. Therefore, it is highly probable that at least some of them will have the same
convictions as the majority of the experimental subjects. Then they will claim that (F2), and
perhaps even (F3), is satisfied. We may, therefore, expect descriptive falsifications of a
decision theory to be accompanied by claims that the theory is invalid from a normative
point of view. Indeed, this is what has often happened.
Decision Under Uncertainty: Prisoner's Dilemma
A common example of decision theory stems from the prisoner's dilemma in which
two individuals are faced with an uncertain decision where the outcome is not only based
on their personal decision, but also on that of the other individual. Since both parties do not
know what actions the other person will take, this results in an uncertain decision
framework. While mathematics and statistical models determine what the optimal decision
should be, psychology and philosophy introduce factors of human behaviors to suggest the
most likely outcome.
Probability and Decision Making - Value and Utility
The future is uncertain, but we have to make decisions every day that have an effect
on our prospects, financial and otherwise. Faced with uncertainty, we do not merely throw
up our hands and guess randomly about what to do; instead, we assess the potential risks
and benefits of a variety of options, and choose to act in a way that maximizes the
probability of a beneficial outcome. Things won’t always turn out for the best, but we have
to try to increase the chances that they will. To do so, we use our knowledge—or at least
our best estimates—of the probabilities of future events to guide our decisions.
The process of decision-making in the face of uncertainty is most clearly illustrated
with examples involving financial decisions. When we make a financial investment, or—
what’s effectively though not legally the same thing—a wager, we’re putting money at risk
with the hope that it will pay off in a larger sum of money in the future. We need a way of
deciding whether such bets are good ones or not. Of course, we can evaluate our financial
decisions in hindsight, and deem the winning bets good choices and the losing one’s bad
choices, but that’s not a fair standard. The question is, knowing what we knew at the time
we made our decision, did we make the choice that maximized the probability of a
profitable outcome, even if the profit was not guaranteed?
To evaluate the soundness of a wager or investment, then, we need to look not at its
worth after the fact—its final value, we might say—but rather at the value we can
reasonably expect it to have in the future, based on what we know at the time the decision
is made. We’ll call this the expected value (sometimes called expectation value). To
calculate the expected value of a wager or investment, we must take into consideration:
1. the various possible ways in which the future might turn out that are relevant to our
bet,
2. the value of our investment in those various circumstances, and
3. the probabilities that these various circumstances will come to pass.
The expected value is a weighted average of the values in the different circumstances; it is
weighted by the probabilities of each circumstance. Here is how we calculate expected
value (EV):
EV = P(O1) x V(O1) + P(O2) x V(O2) + ... + P(On) x V(On)
This formula is a sum; each term in the sum is the product of a probability and a value.
Where: ‘O1, O2, ..., On’ refer to all the possible future outcomes that are relevant to
our bet.
P(Ox) is the probability that outcome #x will come to pass. V(Ox) is the value
of our investment should outcome #x come to pass.
Example 1: You and your roommate are bored, so you decide to play a game; you’ll each put
up a dollar, then flip a coin; if it comes up heads, you win all the money; if it comes up tails,
she does. (In this and what follows, I am indebted to Copi and Cohen’s presentation for
inspiration.) What’s the expected value of your $1 bet?
First, we need to consider which possible future circumstances are relevant to your bet’s
value. Clearly, there are two: the coin comes up heads, and the coin comes up tails.
There are two outcomes in our formula: O1 = heads, O2 = tails.
The probability of each of these is 1/2.
We must also consider the value of your investment in each of these circumstances.
If heads come up, the value is $2—you keep your dollar and snag hers, too.
If tails come up, the value is $0—you look on in horror as she pockets both bills.
(Note: value is different from profit. You make a profit of $1 if heads come up, and you
suffer a loss of $1 if tails do—or your profit is -$1. Value is how much money you’re holding
at the end.)
Plugging the numbers into the formula, we get the expected value:
EV = P(heads) x V(heads) + P(tails) x V(tails) = 1/2 x $2 + 1/2 x $0 = $1
The expected value of your $1 bet is $1. You invested a dollar with the expectation of a
dollar in return. This is neither a good nor a bad bet. A good bet is one for which the
expected value is greater than the amount invested; a bad bet is one for which it’s less.
Example 2: A normal six-sided die. Once you roll the die, it has an equal one-sixth chance of
landing on one, two, three, four, five, or six. Given this information, the calculation is
straightforward:
If you were to roll a six-sided die

an infinite amount of times, you see the average value equals 3.5.
What is a decision tree?
In its simplest form, a decision tree is a type of flowchart that shows a clear pathway
to a decision. In terms of data analytics, it is a type of algorithm that includes conditional
‘control’ statements to classify data. A decision tree starts at a single point (or ‘node’) which
then branches (or ‘splits’) in two or more directions. Each branch offers different possible
outcomes, incorporating a variety of decisions and chance events until a final outcome is
achieved. When shown visually, their appearance is tree-like…hence the name!
Decision trees are extremely useful for data analytics and machine learning because
they break down complex data into more manageable parts. They’re often used in these
fields for prediction analysis, data classification, and regression. Don’t worry if this all
sounds a bit abstract—we’ll provide some examples below to help clear things up. First
though, let’s look at the different aspects that make up a decision tree.
Decision Tree Uses
A decision tree can be used in either a predictive manner or a descriptive manner. In
either instance they are constructed the same way and are always used to visualize all
possible outcomes and decision points that occur chronologically. Decision trees are most
commonly used in the financial world for areas such as loan approval, portfolio
management, and spending. A decision tree can also be helpful when examining the viability
of a new product or defining a new market for an existing product.
Advantages of decision trees
 Good for interpreting data in a highly visual way.
 Good for handling a combination of numerical and non-numerical data.
 Easy to define rules, e.g. ‘yes, no, if, then, else…’
 Requires minimal preparation or data cleaning before use.
 Great way to choose between best, worst, and likely case scenarios.
 Can be easily combined with other decision-making techniques.
Disadvantages of decision trees
 Overfitting (where a model interprets meaning from irrelevant data) can become a
problem if a decision tree’s design is too complex.
 They are not well-suited to continuous variables (i.e. variables which can have more
than one value, or a spectrum of values).
 In predictive analysis, calculations can quickly grow cumbersome, especially when a
decision path includes many chance variables.
 When using an imbalanced dataset (i.e. where one class of data dominates over
another) it is easy for outcomes to be biased in favor of the dominant class.
 Generally, decision trees provide lower prediction accuracy compared to other
predictive algorithms.
What are the different parts of a decision tree?
Decision trees can deal with complex data, which is part of what makes them useful.
However, this doesn’t mean that they are difficult to understand. At their core, all decision
trees ultimately consist of just three key parts, or ‘nodes’:
 Decision nodes: Representing a decision (typically shown with a square)
 Chance nodes: Representing probability or uncertainty (typically denoted by a circle)
 End nodes: Representing an outcome (typically shown with a triangle)
Connecting these different nodes are what we call branches. Nodes and branches can be
used over and over again in any number of combinations to create trees of various
complexity. Let’s see how these parts look before we add any data.
Seven Tips for Creating a Decision Tree
Here are some best practice tips for creating a decision tree diagram:
 Start the tree. Draw a rectangle near the left edge of the page to represent the first
node. In this rectangle, write the first question, main idea, or criterion that will lead
to a decision.
 Add branches. For every possible alternative draw a separate line that begins at the
node and moves away toward the right of the page. Using a loan approval process as
an example, the first node may have been "Income", and the associated branches
might be <$50K, $51K - $100K, >$101K.
 Add leaves. The bulk of the decision tree will be leaf nodes. At the end of each
branch add a leaf node. Fill each of these leaf nodes with another question or
criterion.
 Add more branches. Repeat the process of adding a branch for each possible
alternative leading from a leaf. Label each branch just as before.
 Complete the decision tree. Continue adding leaves and branches until every
question or criterion has been resolved and an outcome has been reached.
 Terminate a branch. Continue adding leaves and branches until every question or
criterion has been resolved and an outcome has been reached.
 Verify accuracy. Consult with all stakeholders to verify accuracy.
Steps in Decision Tree Analysis
Following steps simplify the interpretation process of a decision tree:
1. The first step is understanding and specifying the problem area for which decision
making is required.
2. The second step is interpreting and chalking out all possible solutions to the
particular issue as well as their consequences.
3. The third step is presenting the variables on a decision tree along with its respective
probability values.
4. The fourth step is finding out the outcomes of all the variables and specifying it in the
decision tree.
5. The last step is highly crucial and backs the overall analysis of this process. It involves
calculating the EMV values for all the chance nodes or options, to figure out the
solution which provides the highest expected value.
How to Construct a Decision Tree
A decision tree is a branched flowchart showing multiple pathways for potential
decisions and outcomes. The tree starts with what is called a decision node, which signifies
that a decision must be made.
An example of a decision tree
From the decision node, a branch is created for each of the alternative choices under
consideration. The initial decision might lead to another decision, in which case a new
decision node is created and new branches are added to show each alternative pathway for
the new decision. The result is a series of decision pathways. The flowchart might include
only one or two decisions with only one or two alternatives, or it can become a complex
sequence of many decisions with many alternatives at each node.
Along the decision pathway, there is usually some point at which a decision leads to
an uncertain outcome. That is, a decision could result in multiple possible outcomes, so an
uncertainty node is added to the tree at that point. Branches come from that uncertainty
node showing the different possible outcomes.
Eventually, each pathway reaches a final outcome. The decision tree, then, is a
combination of decision nodes, uncertainty nodes, branches coming from each of these
nodes, and final outcomes as the result of the pathways.
Decision Tree Analysis Example
To enlighten upon the decision tree analysis, let us illustrate a business situation.
ABC Ltd. is a company manufacturing skincare product. It was found that the business is at
the maturity stage, demanding some change. After rigorous research, management came up
with the following decision tree:
In the above decision tree, we can easily make out that the company can expand its
existing unit or innovate a new product, i.e., shower gel or make no changes.
Given below is the evaluation of each of these alternatives:
Expansion of Business Unit:

If the company invests in the development of its business unit, there can be two
possibilities, i.e.:
40% possibility that the market share will hike, increasing the overall profitability of
the company by ₹2500000;
60% possibility that the competitors would take over the market share and the
company may incur a loss of ₹800000.
To find out the viability of this option, let us compute its EMV (Expected Monetary Value):
New Product Line of Shower Gel:
If the organization go for new product development, there can be following two
possibilities:
50% chances are that the project would be successful and yield ₹1800000 as profit;
50% possibility of failure persists, leading to a loss of ₹800000.
To determine the profitability of this idea, let us evaluate its EMV:
Do Nothing:
If the company does not take any step, still there can be two outcomes, discussed below:
40% chances are there that yet, the organization can attract new customers,
generating a profit of ₹1000000;
60% chances of failure are there due to the new competitors, incurring a loss of
₹400000.
Given below is the EMV in such circumstances:

Interpretation
From the above evaluation, we can easily make out that the option of a new product
line has the highest EMV. Therefore, we can say that the company can avail this opportunity
to make the highest gain by ensuring the best possible use of its resources.
A very fast intro to decision theory
There are 4 basic elements in decision theory: acts, events, outcomes, and payoffs. Acts are
the actions being considered by the agent -in the example elow, taking the raincoat or not;
events are occurrences taking place outside the control of the agent (rain or lack thereof);
outcomes are the result of the occurrence (or lack of it) of acts and events (staying dry or
not; being burdened by the raincoat or not); payoffs are the values the decision maker is
placing on the occurrences (for example, how much being free of the nuisance of carrying
an raincoat is worth to one). Payoffs can be positive (staying dry) or negative (the raincoat
nuisance). It is often useful to represent a decision problem by a tree.

Here a square indicates a node in the tree where a decision is made and a circle
where events take place. The tree does not contain payoffs yet, but they can easily be
placed by the outcomes.
In general, we can note two things. First, the nature of the payoffs depends on
one’s objectives. If one is interested only in making money, then payoffs are best accounted
for in terms of money. However, if one is interested in, say, safety, then the payoffs are
best accounted for in terms of risk of accident, for example. If any numerical approach is
possible when disparate objectives are involved, there must be some universal measurable
quantity making them comparable. (In fact, utility, of which more later, is such a
quantity). Second, decision making trees can become unmanageable very fast if one tries
to account for too many possibilities. For example, it would be physically impossible to
account for all the possibilities involved in the decision of which 50 out of 200 gadgets should be
intensively marketed, as the number of possible combinations, 200!/(50! ∙ 150!) is simply
astronomical. Hence one must use good judgment in limiting the options considered; this is
potentially problematic, as one may unwittingly fail to consider a possible action which would
produce very good outcomes.
Decision Criteria
How one uses a decision tree or a decision matrix depends on the decision criteria one
adopts. Consider the following payoff matrix where acts are rows, events columns, and the
resulting squares contain the payoffs (outcomes are not represented to avoid clutter). So,
suppose that we are considering which widget out of 3 to produce and our goal is making
money.

EVENTS
Good sales Bad sales
ACTS
Produce A +$5000 -$1000
Produce B +$10000 -$3000

Produce C +$3000 -$500
Here producing B is obviously the best option if things go well, while producing C is
the best option if things go badly, as losing $500 is the best of the worst payoffs. The
decision criterion telling us to choose C is called “Maximin”. Obviously maximin is a rather
pessimistic strategy, and for this reason it is controversial. However, if the stakes are very
high (for example, suppose that if I lose more than $500 I will be forever ruined), maximin
seems a reasonable option. The application of maximin in the original position has played
an important role in Rawls’ A Theory of Justice, the most important work in political
philosophy in the last decades. Other decision criteria in cases of uncertainty
are maximax, minimax of regret, and the appeal to subjective probabilities through
the Principle of Indifference. Unfortunately, none of these principles is always viable.
However, when the probabilities of events are available (that is, in decision under risk) and
the agent is indifferent to risk, as when the payoffs involved are significant but not too
significant, the criterion usually put forth in decision theory is that of the expected
maximum payoff (EMP), the counterpart of the principle in gambling enjoining us to choose
the bet with the greatest expected value. So, suppose that we could provide the relevant
probabilities, as in the following matrix:

EVENTS
Good sales Bad sales Payoff Expected
payoff
ACTS
Produce A +$5000 x 80% = -$1000 x 20% = - +$4000 +$3800
Pr(good sales)= 4000 200
80%
Pr(bad sales)
=20%
Produce B +$10000 x 60% = -$3000 x 40% = - +$7000 +$4800
Pr(good sales)= 6000 1200
60%
Pr(bad sales)
=40%
Produce C +$3000 x 50% = -$500 x 50% = - +$2500 +$1250
Pr(good sales)= $1500 $250
50%
Pr(bad sales)
=50%

Then, EMP would tell us to produce B, as the expected payoff is the greatest. Most
business decisions fall into this category. For example, if a company makes dozens of
decisions with comparable payoffs every day, then EMP is the best business strategy, as it is
for a casino.
Decision trees and backward induction
Typically, more than one decision is involved in decision making, in which case it is
best to use a tree instead of a matrix. For example, consider the following situation, in
which no probabilities are involved.
You have arrived at a fork in the road on you way home.
If you go left, you’ll have to go through a very sunny patch in the mid of a very hot
day. However, this will also allow you to admire a beautiful bloom of wildflowers growing
by the side of the path. You shall then arrive at another fork. If you take another left, you
will have to go by a neighbor’s house with a very unpleasant guard dog that will growl at
you from the other side of the fence. By contrast, if you go right at the second fork, you’ll
go by a very noisy and dusty part of the road. Whichever of the two you take, you shall get
home quickly.
If you go right at the first fork, you’ll go through the woods, which are very cool this
time of the year. However, there will be little to admire until you get to another fork in the
road. If you go left at this fork, you will see some beautiful meadows; unfortunately, it will
take you longer to reach your home, a bad thing since you are in a bit of a hurry. If you go
right, you shall get home in good time.
Suppose that you assign the following utilities:
Getting hot: -10; seeing the wildflowers: +15; being growled at: -3; pleasant coolness: +10;
being a bit late getting home: -5; taking a noisy and dusty road: -2; seeing the nice
meadows: +4; getting home in good time: +2.
We can construct a decision tree.

Decision trees can be used by applying backwards induction. The idea is that in
order to determine what to do at a (the decision at the first fork), one needs to decide what
one would do at b and c. In other words, the tree is analyzed form the right (from the
outcomes) to the left (to the earlier decisions). So, at b, one would take the right path
because it leads to outcome O2 with utility is +5 while the utility of the left path leading to
O1 is +4. We can represent this choice by pruning the left path, that is, by placing an ‘x’ on
it. By the same token, at c one would choose to go right, and therefore we may place an ‘x’
over the left option. We are now left with a simplified tree at a: going left will have utility
+5, while going right will have utility +12. Hence, we should go right twice.
Adding probabilities
The previous example did not involve probabilities. However, introducing them is
not much of a problem, as the following example shows.
You are about to produce a new garment C and must determine whether to merchandise it
only nationally (N) or internationally as well (I).
If you choose N and sales are good (G), then you’ll make 4, and if they are bad (B)
you’ll lose 1. (All payoffs are in millions). You believe that the probability of good national
sales is .8 and that of bad national sales is .2. You must also decide whether to produce and
sell a matching scarf S. If S’s sales are good, you’ll make an additional 2, and if they are bad,
you’ll lose an additional 1. You think that if C sells well the probability that S’s sales are
good is .9 and the probability that S sells badly is .1. By contrast, if C sells badly, the
probability that S’s sales are good is .4 and the probability that S sells badly is .6.
If you choose I and sales are good you’ll make 6, and if they are bad, you’ll lose 2. You
believe that the probability of good sales is .7 and that of bad sales is .3. As in the other
case, you must decide about S. If S’s sales are good, you’ll make an additional 3, and if they
are bad, you’ll lose an additional 2. You think that if C sells well the probability that S’s sales
are good is .8 and the probability that S sells badly is .2. By contrast, if C sells badly, the
probability that S’s sales are good is .4 and the probability that S sells badly is .6.
Because of the seasonal nature of your business, you need to decide now, a few months in
advance, what to do.
The tree below diagrams the decision. Obviously, the first decision is whether to
choose N or I. Let’s follow the subtree tree arising from N, as that stemming from, I am
analogous. Upon deciding N, there is an act of nature, hence the circle with G (good sales)
and B (bad sales). After G, we write .8, which is the probability of good sales if you choose
N. Your second decision regards S, which is represented by the square with the two
branches S (produce and sell S) and ~S (don’t produce S). If you choose S, then the sales can
be good or bad (an act of nature) represented by the circle and the relevant probabilities; if
you choose ~S, then you are left with the gains or losses from selling C nationally. The
payoff is easy to calculate. Consider the uppermost branch, representing your choice of N,
good sales of C, your choice of S and good sales of S. Since C sells well, you make 4 and
since S sells well, you make 2, so that the total payoff for this branch is 4 + 2 = 6. The second
topmost branch has a payoff of 3 because although C sells well (you gain 4), S does not (you
lose 1).
The tree is analyzed by backward induction. Starting at the top, the last decision you
make is whether to choose S or ~S. The expected payoff of choosing S is 5.7, which is 6
times .9 plus 3 times .1, exactly as if we were considering a bet in which you get 6 with
probability .9 and 3 with probability .1. Since 5.7 > 4, you clip ~S, which simply means that
the best decision given N and good sales of C is to choose S. Since 5.7 is the best expected
payoff available, we write 5.7 next to the relevant square. By contrast, given N and bad
sales of C, the best option is ~S, with payoff of -1, as the option S has an expected payoff of -
1.2. Hence, S is clipped. We can now calculate the expected payoff of choosing N, namely
5.7 times .8 plus -1 times .2, which is 4.36. We write this figure next to the choice N. The
lower half of the tree is analyzed analogously, resulting in an expected payoff of 5 for
I. Since 5 > 4.36, N is clipped and I is the best choice, that with the highest expected
payoff.

Revising decision trees in the light of new information
Consider the following scenario. You have to choose which of two widgets, A and B,
to market. You also believe that the probability of high demand for A is 80% and the
probability of high demand for B is 30%. (High and low demands are the only two
alternatives). However, while you get only $3 for every sold A, you get $5 for every sold
B. Suppose that if an item is in high demand you shall sell 10,000 of them and with low
demand only 4000. Item A cost $1 to produce, while item B costs $2. Moreover, time
constraints due to the holiday season compel you to produce 10,000 items of whichever of
the two widgets you decide to market. What should you do?
We already know how to construct the decision tree listing the outcomes and the
payoffs in thousands. For example, since 10,000 items must be produced no matter what
the sales will be, the production cost of A is 10. If A sells well, all 10,000 will be sold, with a
gain of 20. Hence, the payoff will be 20-10=10. If demand is low, only 4,000 will be sold,
with a gain of 8. Hence, the payoff will be 8-10=-2.
By regressive induction, we see that MA has an expected payoff of 7.6 and MB of
3.2, which means that we should choose MA.

Suppose, however, that you submit A and B to a buyer panel; the reliability of the
panel is given by the following figures: in the past, out of 100 high demand items, the panel
was positive (P) about 70 and negative (N) about 30; out of 100 low demand items, the
panel was positive about 20 and negative about 80. (Note that high or low demand is built
into the market, and therefore is the conditioning factor in conditional probability). In short,
Pr(P|HD)=70%; Pr(N|HD)=30%; Pr(N|LD)=80%; Pr(P|LD)=20%.
To make use of this new information, we use Bayes’ theorem to determine the posterior
probabilities of HD and LD, that is, their probabilities given the panel’s results. The
relevant probability, not decision, tree is

Hence, by Bayes’ theorem (with figures adjusted to 2 decimals):

Pr(HDA|P) = .93; Pr(LDA|P) = .07
Pr(HDA|N) = .6; Pr(LDA|N) = .4.
By an analogous procedure, one obtains
Pr(HDB|P) = .6; Pr(LDB|P) = .4
Pr(HDB|N) = .14; Pr(LDB|N) = .86.
At this point we can rewrite the decision tree incorporating the new posterior
probabilities:

From now on it is just a matter of employing backward induction to determine which
widget to market. The expected payoff of marketing A if the panel has a positive reaction is
+$7.9; if the panel has a negative reaction the expected payoff drops to +$5.2. With respect
to B, if the panel has a positive reaction, the expected payoff is $10.4; if the panel has a
negative reaction, it becomes -$.64. The expected value of marketing A is 6.8; that of
marketing B is 3.2. Hence, one should market A. Since the expected payoff difference
between MA and MB is 3.6, one should pay less than 3.6 for the panel; in other words, given
the data of the problem, the extra information is not worth more than 3.6.
One can easily see that decision trees could get complex and difficult to
construct. For example, getting the information may take time when time is of the essence
(perhaps, A and B are holiday season widgets). Moreover, in our examples we came up with
the relevant probabilities by fiat. In real situations, of course, coming up with the correct
probabilities can be very difficult, especially when appealing to history or statistical surveys
does not help. However, such cases are best left to a decision theory course.
Solution:

Qba 221-Quantitative Business Analysis

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Qba 221-Quantitative Business Analysis

Uploaded by

Copyright:

Available Formats

MODULE IN QBA 221 (QUANTITATIVE BUSINESS ANALYSIS) – BSAB

INSTRUCTION TO THE USER:

TYPES OF SYSTEMS OF LINEAR EQUATIONS:

B. Dependent. A dependent solution contains an infinite number of solutions and the

C. Consistent. A consistent contains only one solution, and

The solution is (1, -5)

Substituting the value of x

Using Cramer’s Rule to Solve a System of Two Equations in Two Variables

Solving for y gives

 Dy: determinant of the numerator in the solution of y

The solution using Cramer’s Rule is given as:

The solution is (2, -3).

Now, replace y with 0 to determine the y-intercept.

The x-intercept is (0, 4) and the y-intercept is (4, 0).

y = x - 1 y = -2x - 7

Symbolically, x = y + 1 2x + y = -7 inequalities are

B. Addition and Subtraction Property

( 12 ) ( 2 x ) >( 12 ) ( 12) Multiply both sides by ½.

GRAPHING LINEAR INEQUALITY

SYSTEM OF LINEAR INEQUALITIES

How to Solve Systems of Linear Inequalities?

LESSON 2: LINEAR PROGRAMMING: GRAPHICAL METHOD

Every student in business administration has a secret dream of becoming an executive or

 Solving Linear Programming Problem Graphically

A linear programming problem in two unknowns x and y is one in which we are to

Two types of constraints:

a. Structural constraint – is a limit on the availability of resources; it is also referred as

The Linear programming model:

D. Solving Linear Programming Maximization Problem

Apply exponential smoothing with a smoothing constant of 0.7 to generate a

 Linear Programming Problem Involving Mixed Constraints

Solve the following:

3.) Solve the linear programming problem:

LESSON 3: LINEAR PROGRAMMING: THE SIMPLEX METHOD

To obtain a zero in the entry first above the

To obtain a zero in the element below the pivot,

We can restate the solution associated with this matrix as x1=8, x2=0, y1=4, y2=0 and z=320.

We make the pivot element 1 by multiplying row 1 by 2, and we get

We no longer have negative entries in the

The matrix reads x1=4, x2=8 and z=400.

The completely transformed minimization model can now be summarized as

We restate the solution as follows:

1. Maximize P = 40x1 + 30x2

Subject to: x1 + 2x2 ≤ 16

3. Find solution using Simplex (Big M) method

LESSON 4: SPECIAL PURPOSE ALGORITHMS PROGRAMMING

Transportation and Assignment Models

Basic structure of transportation problem:

The following is the initial basic feasible solution:

– For U-V method the

Optimization of the solution using U-V Method:

This means that programmer 1 is assigned programme C, programmer 2 is assigned

What is Decision Theory?

Normative Decision Theory Optimal Decision Theory

The theory centers around the ideal

"What should the person making the

Example of Expected Value (Multiple Events)

New Product Line of Shower Gel:

Given below is the EMV in such circumstances:

What is decision theory?

Decision Theory: Bayesian

If you were to roll a six-sided die