Sensitivity Analysis......................................................................................................

2
Definition:...........................................................................................................................2
Introduction:.......................................................................................................................2
Impacts of Sensitivity Analysis:.........................................................................................2
Uses of Sensitivity Analysis:..............................................................................................3
Effect of parameter and characteristics of the model on the optimality:.......................5
Range of Optimality:..........................................................................................................5
Example:.............................................................................................................................5
Example:.............................................................................................................................7
Cowhide...........................................................................................................................7
Production Time..............................................................................................................7
Production Limit..............................................................................................................8
Non-negativity of Decision Variables..............................................................................8
The Mathematical Model.................................................................................................8
Decision Analysis..........................................................................................................9
Definition:...........................................................................................................................9
Introduction:.......................................................................................................................9
Methodology:....................................................................................................................10
Decision-making under uncertainty (or risk) Seven Steps:...........................................10
Different Criterion:..........................................................................................................11
Approaches for Decision under Risk:.............................................................................11
Approaches for Decision under uncertainty:.................................................................12
Optimistic criterion........................................................................................................12
Minimax (Regret) criterion............................................................................................12
Laplace criterion............................................................................................................12
Savage Minimax Regret criterion...................................................................................13
Decision Tree with Examples:.........................................................................................13
Decision tree example 1995 UG exam:..........................................................................13
Solution.............................................................................................................................15

Sensitivity Analysis

Definition:
“Investigation into how projected performance varies along with changes in the
key assumptions on which the projections are based.”
Introduction:
Sensitivity analysis is used to determine how “sensitive” a model is to changes in the
value of the parameters of the model and to changes in the structure of the model.
Sensitivity analysis helps to build confidence in the model by studying the
uncertainties that are often associated with parameters in models. Many parameters in
system dynamics models represent quantities that are very difficult, or even
impossible to measure to a great deal of accuracy in the real world. Also, some
parameter values change in the real world. Therefore, when building a system
dynamics model, the modeler is usually at least somewhat uncertain about the
parameter values he chooses and must use estimates.

Impacts of Sensitivity Analysis:

Sensitivity analysis can be useful to computer modelers for a range of purposes
including,
Support decision making or the development of recommendations for decision makers
(e.g. testing the robustness of a result);
Enhancing communication from modelers to decision makers (e.g. by making
recommendations more credible, understandable, compelling or persuasive);
Increased understanding or quantification of the system (e.g. understanding
relationships between input and output variables); and
Model development (e.g. searching for errors in the model).

Uses of Sensitivity Analysis:

There is a very wide range of uses to which sensitivity analysis is put. The uses are
grouped into four main categories: decision making or development of
recommendations for decision makers, communication, increased understanding or
quantification of the system, and model development.

Decision Making or Development of Recommendations for Decision

1.
Makers

2
1.
Testing the robustness of an optimal solution.
1
1. Identifying critical values, thresholds or break-even values where the
2 optimal strategy changes.
1.
Identifying sensitive or important variables.
3
1.
Investigating sub-optimal solutions.
4
1.
Developing flexible recommendations, which depend on circumstances.
5
1.
Comparing the values of simple and complex decision strategies.
6
1.
Assessing the "riskiness" of a strategy or scenario.
7
   
2. Communication
2. Making recommendations more credible, understandable, compelling or
1 persuasive.
2.
Allowing decision makers to select assumptions.
2
2.
Conveying lack of commitment to any single strategy.
3
   
3. Increased Understanding or Quantification of the System
3.
Estimating relationships between input and output variables.
1
3.
Understanding relationships between input and output variables.
2
3.
Developing hypotheses for testing
3
   
4. Model Development
4.
Testing the model for validity or accuracy.
1
4.
Searching for errors in the model.
2
4. Simplifying the model.

3
 3
4.
Calibrating the model.
4
4.
Coping with poor or missing data.
5
4.
Prioritizing acquisition of information.
6
 
In all models, parameters are more-or-less uncertain. The modeler is likely to be
unsure of their current values and to be even more uncertain about their future values.
This applies to things such as prices, costs, productivity, and technology. Uncertainty
is one of the primary reasons why sensitivity analysis is helpful in making decisions
or recommendations. If parameters are uncertain, sensitivity analysis can give
information such as:
a. How robust the optimal solution is in the face of different parameter values (use
1.1)
b. Under what circumstances the optimal solution would change (uses 1.2, 1.3, 1.5)
c. How the optimal solution changes in different circumstances (use 3.1)
d. How much worse off would the decision makers be if they ignored the changed
circumstances and stayed with the original optimal strategy or some other strategy
(uses 1.4, 1.6)
This information is extremely valuable in making a decision or recommendation. If
the optimal strategy is robust (insensitive to changes in parameters), this allows
confidence in implementing or recommending it. On the other hand if it is not robust,
sensitivity analysis can be used to indicate how important it is to make the changes to
management suggested by the changing optimal solution. Perhaps the base-case
solution is only slightly sub-optimal in the plausible range of circumstances, so that it
is reasonable to adopt it anyway. Even if the levels of variables in the optimal solution
are changed dramatically by a higher or lower parameter value, one should examine
the difference in profit (or another relevant objective) between these solutions and the
base-case solution. If the objective is hardly affected by these changes in
management, a decision maker may be willing to bear the small cost of not altering
the strategy for the sake of simplicity.

4
Effect of parameter and characteristics of the model on the optimality:
In many conditions, the parameter and characteristics of a linear programming model
may changeover a period of time. Also, the analyst may be interested to know the
effect of changing the parameters and characteristics of the model on the optimality.
This kind of sensitivity analysis can be carried out in the following ways:
Making changes in the right-hand side constants of the constraints
Making changes in the objective function coefficients
Adding a new constraint
Adding a new variable

Range of Optimality:
The range of optimality for each coefficient provides the range of values over which
the current solution will remain optimal. Managers should focus on those objective
coefficients that have a narrow range of optimality and coefficients near the endpoints
of the range.
Graphically, the limits of a range of optimality are found by changing the slope of the
objective function line within the limits of the slopes of the binding constraint lines.
Binding constraints are the constraints that intersect to form the optimal point.
We will focus on using the computer output to analyze the range of optimality for
objective function coefficients.

Example:
Before identifying the range of optimality.

5
Now as discuss that for in graphical solution we can find the range of optimality by
changing the slope of the objective function such that it nearly lies within the
constraints line.

By changing coefficients in the objective function you are altering its slope. You can
verify this by simply changing a coefficient and graphing the new objective function.

6
Example:
Wilson Problem: Wilson Manufacturing produces both baseballs and softballs, which
it wholesales to vendors around the country. Its facilities permit the manufacture of a
maximum of 500 dozen baseballs and a maximum of 500 dozen softballs each day.
The cowhide covers for each ball are cut from the same processed cowhide sheets.
Each dozen baseballs require five square feet of cowhide (including waste), whereas,
one dozen softballs require six square feet of cowhide (including waste). Wilson has
3600 square feet of cowhide sheets available each day. Production of baseballs and
softballs includes making the inside core, cutting and sewing the cover, and
packaging. It takes about one minute to manufacture a dozen baseballs and two
minutes to manufacture dozen softballs. A total of 960 minutes is available for
production daily.

Formulate a set of linear constraints that characterize the production process at Wilson
Manufacturing.
Decision Variables
X1= number of dozen baseballs produced daily
X2= number of dozen softballs produced daily
Constraints
In addition to non-negativity constraints (i.e., the implied constraints) for the decision
variables, there are three functional constraints.
The use of cowhide.
The daily limit for production time.
The maximum production limit of total units.

Cowhide
The total amount of cowhide used daily cannot exceed the amount of cowhide
available daily
5X1 + 6X2 £ 3600

Production Time
The amount of production minutes used daily cannot exceed the total number of
production minute’s available daily
X1 + 2X2 £ 960

7
Production Limit
The total number of dozen units produced daily cannot exceed the marketing limits
X1 £ 500
X2 £ 500

Non-negativity of Decision Variables

Negative Production of baseballs and softballs is impossible. Thus,
X1, X2 3 0

The Mathematical Model

Max 7X1+ 10X2 (Objective Function)

Subject to:
5X1 + 6X2 £ 3600 (Cowhide)
X1 + 2X2 £ 960 (Production time)
X1 £ 500 (Production limit of baseballs)
X2 £ 500 (Production limit of softballs)
X1, X2 3 0 (Non-negativity)

8
Decision Analysis

Definition:
Decision analysis is a logical and systematic way to address a wide variety of
problems involving decision-making in an uncertain environment. We introduce the
method of decision analysis and the analytical model of constructing and solving a
decision tree with the following prototypical decision problem.

Introduction:
The term decision analysis was coined in 1964 by Ronald A. Howard, who since then,
as a professor at Stanford University, has been instrumental in developing much of the
practice and professional application of DA.
Graphical representation of decision analysis problems commonly use influence
diagrams and decision trees. Both of these tools represent the alternatives available to
the decision maker, the uncertainty they face, and evaluation measures representing
how well they achieve their objectives in the final outcome. Uncertainties are
represented through probabilities and probability distributions. The decision maker's
attitude to risk is represented by utility functions and their attitude to trade-offs
between conflicting objectives can be made using multi-attribute value functions or
multi-attribute utility functions (if there is risk involved). In some cases, utility
functions can be replaced by the probability of achieving uncertain aspiration levels.
Decision analysis advocates choosing that decision whose consequences have the
maximum expected utility (or which maximize the probability of achieving the
uncertain aspiration level). Such decision analytic methods are used in a wide variety
of fields, including business  healthcare research and management, energy
exploration, litigation and dispute resolution, etc. An applied branch of decision
theory.
Decision analysis offers individuals and organizations a methodology for making
decisions; it also offers techniques for modeling decision problems mathematically
and finding optimal decisions numerically. Decision models have the capacity for
accepting and quantifying human subjective inputs: judgments of experts and
preferences of decision-makers. Implementation of models can take the form of
simple paper-and-pencil procedures or sophisticated computer programs known as
decision aids or decision systems.

9
Methodology:
The methodology is rooted in postulates of rationality—a set of properties which
preferences of rational individuals must satisfy. One such property is transitivity: if an
individual prefers action ato action b and action b to action c, he or she should
prefer a to c. From the rationality postulates, principles of decision-making are
derived mathematically. The principles prescribe how decisions ought to be made, if
one wishes to be rational. In that sense, decision analysis is normative.
The methodology is broad and must always be adapted to the problem at hand. An
illustrative adaptation to a class of problems known as decision-making under
uncertainty (or risk) is outlined in the illustration and consists of seven steps:

Decision-making under uncertainty (or risk) Seven Steps:

The problem is structured by identifying feasible actions, one of which must be
decided upon; possible events, one of which occurs thereafter; and outcomes, each of
which results from a combination of decision and event. Problem structuring can be
facilitated by displays such as decision trees and decision matrices.
At the time of decision-making, the event that will actually occur cannot be predicted
perfectly. The degree of certainty about the occurrence of an event, given all
information at hand, is quantified in terms of the probability of the event.
Preferences are personal: the same outcome may elicit different degrees of desirability
from different individuals. The subjective value that a decision-maker attaches to an
outcome is quantified and termed the utility of outcome.
The preceding steps conform to the principle of decomposition: probabilities of events
and utilities of outcomes must be measured independently of one another. They are
next combined in a criterion for evaluating decisions. The utility of a decision is
defined as the expected utility of the outcome. The optimal, or the most preferred,
decision is one with the maximum utility.
The probability encodes the current state of information about a possible event. Often,
additional information can be acquired in the hope of reducing the uncertainty. The
monetary value of such information is computed before purchase and compared with
the cost of information. Thus, one can determine whether or not acquiring information
is economically rational.
The source of information may be a real-world experiment, a laboratory test, a
mathematical model, or the knowledge of an expert. The in formativeness of the

10
source is described in terms of a probabilistic relation between information and event.
This relation, known as the likelihood function, makes it possible to revise the prior
probability of the event and to obtain a posterior probability of the event, conditional
on additional information. The revision is carried out via Bayes' rule.
Given the additional information, prior probabilities can be replaced by posterior
probabilities, and the analysis can be repeated from step 4 onward. Steps 4–6 may be
cycled many times, until the cost of additional information exceeds its value, at which
moment the optimal decision is implemented.
Decision analysis provides a rich set of concepts and techniques to aid the decision
maker in dealing with complex decision problems under uncertainty.
Decision-making can be broadly classified into three broad categories.
 Decision making under certainty
 Decision Making under Risk
 Decision making under uncertainty

Different Criterion:

Approaches for Decision under Risk:

Most of the decisions are made on the basis of some criterion. When there is certainty
or the outcome is sure, decision-making is simpler. When the outcome is not sure,
then different criteria are used. They are for decision making under risk.
Expected Value Criterion
Combined expected value and variance criterion
Known aspiration level criterion
Most likely occurrence criterion

11
Approaches for Decision under uncertainty:

Optimistic criterion
An optimistic decision maker would use the optimistic approach. The decision with
the largest possible payoff is chosen. If the payoff table were in terms of costs, the
decision with the lowest cost would be chosen.

Minimax (Regret) criterion

The minimax regret approach requires the construction of a regret table or an
opportunity loss table. This is done by calculating for each state of nature the
difference between each payoff and the largest payoff for that state of nature. Then,
using this regret table, the maximum regret for each possible decision is listed. The
decision chosen is the one corresponding to the minimum of the maximum regrets.

12
Laplace criterion
Equally likely, also called Laplace, criterion finds decision alternative with highest
average payoff.
First calculate average payoff for every alternative.
Then pick alternative with maximum average payoff.

Savage Minimax Regret criterion

The Minimax Regret criterion focuses on avoiding regrets that may result from
making a non-optimal decision. Although regret is a subjective emotional state, the
assumption is made that it is quantifiable in direct (linear) relation to the rewards of
the payoff matrix.
Regret is defined as the opportunity loss to the decision maker if action alternative Ai

is chosen and state of nature Sj happens to occur. Opportunity loss is the payoff

difference between the best possible outcome under Sj and the actual outcome

resulting from choosing Ai. Formally:

OLij = (row j maximum payoff) - Rij for positive-flow payoffs (profits, income)

OLij = Rij - (row j minimum payoff) for negative-flow payoffs (costs, losses)

where Rij is the reward value (payoff) for column i and row j of the payoff matrix R.

Note that opportunity losses are defined as nonnegative numbers. The best possible
OL is zero (no regret) and the higher the OL value, the greater the regret.

Decision Tree with Examples:

Decision tree example 1995 UG exam:

Your company is considering whether it should tender for two contracts (MS1 and
MS2) on offer from a government department for the supply of certain components.
The company has three options:
Tender for MS1 only; or
Tender for MS2 only; or
Tender for both MS1 and MS2.
If tenders are to be submitted the company will incur additional costs. These costs
will have to be entirely recouped from the contract price. The risk, of course, is that if
a tender is unsuccessful the company will have made a loss.

13
The cost of tendering for contract MS1 only is £50,000. The component supply cost if
the tender is successful would be £18,000.
The cost of tendering for contract MS2 only is £14,000. The component supply cost if
the tender is successful would be £12,000.
The cost of tendering for both contracts MS1 and contract MS2 is £55,000. The
component supply cost if the tender is successful would be £24,000.
For each contract, possible tender prices have been determined. In addition, subjective
assessments have been made of the probability of getting the contract with a particular
tender price as shown below. Note here that the company can only submit one tender
and cannot, for example, submit two tenders (at different prices) for the same
contract.

In the event that the company tenders for both MS1 and MS2 it will either win both
contracts (at the price shown above) or no contract at all.
What do you suggest the company should do and why?
What are the downside and the upside of your suggested course of action?
A consultant has approached your company with an offer that in return for £20,000 in
cash she will ensure that if you tender £60,000 for contract MS2 only your tender is
guaranteed to be successful. Should you accept her offer or not and why?

14
Solution
The decision tree for the problem is shown below.

15
16