Man Sci reviewer for Midterms

Chapter 6: Distribution and network models

Summary
In this chapter we introduced models related to supply chain problems—specifically, transportation
and transshipment problems—as well as assignment, shortest-route, and maximal flow
problems. All of these types of problems belong to the special category of linear programs called
network flow problems. In general, the network model for these problems consists of nodes
representing origins, destinations, and, if necessary, transshipment points in the network system.
Arcs are used to represent the routes for shipment, travel, or flow between the various nodes.
Transportation problems and transshipment problems are commonly encountered when
dealing with supply chains. The general transportation problem has m origins and n destinations.
Given the supply at each origin, the demand at each destination, and unit shipping cost
between each origin and each destination, the transportation model determines the optimal
amounts to ship from each origin to each destination. The transshipment problem is an
extension of the transportation problem involving transfer points referred to as transshipment nodes.
In this more general model, we allow arcs between any pair of nodes in the network.
The assignment problem is a special case of the transportation problem in which all
supply and all demand values are 1. We represent each agent as an origin node and each task
as a destination node. The assignment model determines the minimum cost or maximum
profit assignment of agents to tasks.
The shortest-route problem finds the shortest route or path between two nodes of a
network. Distance, time, and cost are often the criteria used for this model. The shortestroute
problem can be expressed as a transshipment problem with one origin and one destination. By
shipping one unit from the origin to the destination, the solution will determine the
shortest route through the network.
The maximal flow problem can be used to allocate flow to the arcs of the network so
that flow through the network system is maximized. Arc capacities determine the maximum
amount of flow for each arc. With these flow capacity constraints, the maximal flow problem
is expressed as a capacitated transshipment problem.
In the last section of the chapter, we showed how a variation of the transshipment problem
could be used to solve a production and inventory problem. In the chapter appendix we show
how to use Excel to solve three of the distribution and network problems presented in the chapter.
Glossary
Arcs  The lines connecting the nodes in a network.
Assignment problem A network flow problem that often involves the assignment of agents to tasks;
it can be formulated as a linear program and is a special case of the transportation problem.
Capacitated transportation problem A variation of the basic transportation problem in
which some or all of the arcs are subject to capacity restrictions.
Capacitated transshipment problem A variation of the transshipment problem in which some or all
of the arcs are subject to capacity restrictions.
Dummy origin An origin added to a transportation problem to make the total supply equal to the
total demand. The supply assigned to the dummy origin is the difference between the total demand
and the total supply.
Flow capacity The maximum flow FOR an arc of the network. The flow capacity in one direction may
not equal the flow capacity in the reverse direction.
Maximal flow The maximum amount of flow that can ENTER and exit a network system during a
given period of time.
Network A graphical representation of a problem consisting of numbered circles (nodes)
interconnected by a series of lines (arcs); arrowheads on the arcs show the direction of flow.
Transportation, assignment, and transshipment problems are network flow problems.
Nodes The intersection or junction points of a network.
Shortest route Shortest path between two nodes in a network.
Supply chain The set of all interconnected resources involved in producing and distributing a
product.
Transportation problem A network flow problem that often involves minimizing the
cost of shipping goods from a set of origins to a set of destinations; it can be formulated and solved
as a linear program by including a variable for each arc and a constraint for each node.
Transshipment problem An extension of the transportation problem to distribution problems
involving transfer points and possible shipments between any pair of nodes.

Chapter 7: Integer linear programming

Summary
In this chapter we introduced the important extension of linear programming referred to as
integer linear programming. The only difference between the integer linear programming
problems discussed in this chapter and the linear programming problems studied in previous
chapters is that one or more of the variables must be integer. If all variables must be integer,
we have an all-integer linear program. If some, but not necessarily all, variables must be
integer, we have a mixed-integer linear program. Most integer programming applications
involve 0-1, or binary, variables.
Studying integer linear programming is important for two major reasons. First, integer linear
programming may be helpful when fractional values for the variables are not permitted. Rounding a
linear programming solution may not provide an optimal integer solution; methods for
finding optimal integer solutions are needed when the economic consequences of rounding are
significant. A second reason for studying integer linear programming is the increased modeling
flexibility provided through the use of 0-1 variables. We showed how 0-1 variables could be
used to model important managerial considerations in capital budgeting, fixed cost, distribution
system design, bank location, and product design/market share applications.
The number of applications of integer linear programming continues to grow rapidly.
This growth is due in part to the availability of good integer linear programming software
packages. As researchers develop solution procedures capable of solving larger integer linear
programs and as computer speed increases, a continuation of the growth of integer programming
applications is expected.

Glossary
0-1 integer linear program An all-integer or mixed-integer linear program in which the integer
variables are only permitted to assume the values 0 or 1. Also called binary integer program.
All-integer linear program An integer linear program in which all variables are required to be integer.
Capital budgeting problem A 0-1 integer programming problem that involves choosing which
projects or activities provide the best investment return.
Conditional constraint A constraint involving 0-1 variables that does not allow certain variables to
equal 1 unless certain other variables are equal to 1.
Corequisite constraint A constraint requiring that two 0-1 variables be equal. Thus, they are both in
or out of solution together.
Distribution system design problem A mixed-integer linear program in which the binary integer
variables usually represent sites selected for warehouses or plants and continuous variables
represent the amount shipped over arcs in the distribution network.
Fixed cost problem A 0-1 mixed-integer programming problem in which the binary variables
represent whether an activity, such as a production run, is undertaken (variable 5 1) or not (variable
5 0).
Integer linear program A linear program with the additional requirement that one or more of the
variables must be integer.
k out of n alternatives constraint An extension of the multiple-choice constraint. This constraint
requires that the sum of n 0-1 variables equal k.
Location problem A 0-1 integer programming problem in which the objective is to select
the best locations to meet a stated objective. Variations of this problem (see the bank location
problem in Section 7.3) are known as covering problems.
LP Relaxation The linear program that results from dropping the integer requirements for the
variables in an integer linear program.
Mixed-integer linear program An integer linear program in which some, but not necessarily all,
variables are required to be integer.
Multiple-choice constraint A constraint requiring that the sum of two or more 0-1 variables
equal 1. Thus, any feasible solution makes a choice of which variable to set equal to 1.
Mutually exclusive constraint A constraint requiring that the sum of two or more 0-1 variables be
less than or equal to 1. Thus, if one of the variables equals 1, the others must equal 0. However, all
variables could equal 0.
Product design and market share optimization problem Sometimes called the share of choices
problem, it involves choosing a product design that maximizes the number of consumers preferring
it.

Chapter 8: Nonlinear Optimization Models

Summary
In this chapter we introduced nonlinear optimization models. A nonlinear optimization
model is a model with at least one nonlinear term in either a constraint or the objective
function. Because so many processes in business and nature behave in a nonlinear fashion,
allowing nonlinear terms greatly increases the number of important applications that can be
modeled as an optimization problem. Numerous problems in portfolio optimization, pricing,
blending, economics, marketing, facility location, forecasting, and scheduling lend themselves to
nonlinear models.
Unfortunately, nonlinear optimization models are not as easy to solve as linear optimization models,
or even linear integer optimization models. As a rule of thumb, if a problem can
be modeled realistically as a linear or linear integer problem, then it is probably best to do so.
Many nonlinear formulations have local optima that are not globally optimal. Because most
nonlinear optimization codes will terminate with a local optimum, the solution returned by
the code may not be the best solution available. However, as pointed out in this chapter, numerous
important classes of optimization problems, such as the Markowitz portfolio model,
are convex optimization problems. For a convex optimization problem, a local optimum is
also the global optimum. Additionally, the development of nonlinear optimization codes that
do find global optimal solutions is proceeding at a rapid rate.
Glossary
Concave function A function that is bowl-shaped down: For example, the functions f(x) = 25x2 2 5x
and f(x, y) = 2x2 2 11y2 are concave functions.
Convex function A function that is bowl-shaped up: For example, the functions f(x) = x2 2 5x and f(x,
y) 5 x2 15y2 are convex functions.
Global maximum A feasible solution is a global maximum if there are no other feasible points with a
larger objective function value in the entire feasible region. A global maximum is also a local
maximum.
Global minimum A feasible solution is a global minimum if there are no other feasible points with a
smaller objective function value in the entire feasible region. A global minimum is also a local
minimum.
Global optimum A feasible solution is a global optimum if there are no other feasible points with a
better objective function value in the entire feasible region. A global optimum may be either a global
maximum or a global minimum.
Index fund A portfolio of stocks, mutual funds, or other securities that matches as closely as possible
the performance of a broad market index such as the S&P 500.
Local maximum A feasible solution is a local maximum if there are no other feasible solutions with a
larger objective function value in the immediate neighborhood.
Local minimum A feasible solution is a local minimum if there are no other feasible solutions with a
smaller objective function value in the immediate neighborhood.
Local optimum A feasible solution is a local optimum if there are no other feasible solutions with a
better objective function value in the immediate neighborhood. A local optimum may be either a local
maximum or a local minimum.
Markowitz mean-variance portfolio model A portfolio optimization model used to construct a
portfolio that minimizes risk subject to a constraint requiring a minimum level of return.

Chapter 9: Project scheduling:PERT/CPM

Summary
In this chapter we showed how PERT/CPM can be used to plan, schedule, and control a wide
variety of projects. The key to this approach to project scheduling is the development of a
PERT/CPM project network that depicts the activities and their precedence relationships.
From this project network and activity time estimates, the critical path for the network and
the associated critical activities can be identified. In the process, an activity schedule showing the
earliest start and earliest finish times, the latest start and latest finish times, and the slack for each
activity can be identified.
We showed how we can include capabilities for handling variable or uncertain activity
times and how to use this information to provide a probability statement about the chances
the project can be completed in a specified period of time. We introduced crashing as a
procedure for reducing activity times to meet project completion deadlines, and we showed
how a linear programming model can be used to determine the crashing decisions that will
minimize the cost of reducing the project completion time.
Glossary
Activities Specific jobs or tasks that are components of a project. Activities are represented by nodes
in a project network.
Backward pass Part of the PERT/CPM procedure that involves moving backward through the network
to determine the latest start and latest finish times for each activity.
Beta probability distribution A probability distribution used to describe activity times.
Crashing The shortening of activity times by adding resources and hence usually increasing cost.
Critical activities The activities on the critical path.
Critical path The longest path in a project network.
Critical path method (CPM) A network-based project scheduling procedure.
Earliest finish time The earliest time an activity may be completed.
Earliest start time The earliest time an activity may begin.
Expected time The average activity time.
Forward pass Part of the PERT/CPM procedure that involves moving forward through the project
network to determine the earliest start and earliest finish times for each activity.
Immediate predecessors The activities that must be completed immediately prior to the start of a
given activity.
Latest finish time The latest time an activity may be completed without increasing the project
completion time.
Latest start time The latest time an activity may begin without increasing the project completion
time.
Most probable time The most probable activity time under normal conditions.
Optimistic time The minimum activity time if everything progresses ideally.
Path A sequence of connected nodes that leads from the Start node to the Finish node.
Pessimistic time The maximum activity time if substantial delays are encountered.
Program evaluation and review technique (PERT) A network-based project scheduling procedure.
Project network A graphical representation of a project that depicts the activities and shows the
predecessor relationships among the activities.
Slack The length of time an activity can be delayed without affecting the project completion time.

Chapter 10: Inventory Models

Summary
In this chapter we presented some of the approaches used to assist managers in establishing low-cost
inventory policies. We first considered cases for which the demand rate for the
product is constant. In analyzing these inventory systems, total cost models were developed;
these models included ordering costs, holding costs, and, in some cases, backorder costs.
Then minimum cost formulas for the order quantity Q were presented. A reorder point r can
be established by considering the lead-time demand.
In addition, we discussed inventory models for which a deterministic and constant rate
could not be assumed, and thus demand was described by a probability distribution. A critical
issue with these probabilistic inventory models is obtaining a probability distribution that most
realistically approximates the demand distribution. Solution procedures were then presented for
multiperiod models based on either an order-quantity, reorder point, continuous review system or a
replenishment-level, periodic review system.In closing this chapter, we reemphasize that inventory
and inventory systems can be an expensive phase of a firm’s operation. It is important for managers
to be aware of the cost of inventory systems and to make the best possible operating policy decisions
for the inventory system. Inventory models, as presented in this chapter, can help managers to
develop good inventory policies.
Glossary
Backorder The receipt of an order for a product when no units are in inventory. These backorders
are eventually satisfied when a new supply of the product becomes available.
Constant demand rate An assumption of many inventory models that states that the same number of
units are taken from inventory each period of time.
Constant supply rate A situation in which the inventory is built up at a constant rate over a period of
time.
Continuous review inventory system A system in which the inventory position is monitored or
reviewed on a continuous basis so that a new order can be placed as soon as the reorder point is
reached.
Cost of capital The cost a firm incurs to obtain capital for investment. It may be stated as an annual
percentage rate, and it is part of the holding cost associated with maintaining inventory.
Cycle time The length of time between the placing of two consecutive orders.
Deterministic inventory model A model where demand is considered known and not subject to
uncertainty.
Economic order quantity (EOQ) The order quantity that minimizes the annual holding cost plus the
annual ordering cost.
Goodwill cost A cost associated with a backorder, a lost sale, or any form of stock-out or unsatisfied
demand. This cost may be used to reflect the loss of future profits because a customer experienced
an unsatisfied demand.
Holding cost The cost associated with maintaining an inventory investment, including the cost of the
capital investment in the inventory, insurance, taxes, warehouse overhead, and so on. This cost may
be stated as a percentage of the inventory investment or as a cost per unit.
Incremental analysis A method used to determine an optimal order quantity by comparing the cost of
ordering an additional unit with the cost of not ordering an additional unit.
Inventory position The inventory on hand plus the inventory on order.
Lead time The time between the placing of an order and its receipt in the inventory system.
Lead-time demand The number of units demanded during the lead-time period.
Lead-time demand distribution The distribution of demand that occurs during the leadtime period.
Lot size The order quantity in the production inventory model.
Ordering cost The fixed cost (salaries, paper, transportation, etc.) associated with placing an order for
an item.
Periodic review inventory system A system in which the inventory position is checked or reviewed at
predetermined periodic points in time. Reorders are placed only at periodic review points.
Probabilistic inventory model A model where demand is not known exactly; probabilities must be
associated with the possible values for demand.
Quantity discounts Discounts or lower unit costs offered by the manufacturer when a customer
purchases larger quantities of the product.
Reorder point The inventory position at which a new order should be placed.
Safety stock Inventory maintained in order to reduce the number of stock-outs resulting from higher
than-expected demand.
Setup cost The fixed cost (labor, materials, lost production) associated with preparing for a new
production run.
Shortage or stock-out Occurrence when demand cannot be supplied from inventory.
Single-period inventory model An inventory model in which only one order is placed for the product,
and at the end of the period either the item has sold out or a surplus of unsold items will be sold for a
salvage value.

Problem Solving (40 points)

-Assignment Problem
-PERT
-CPM
-EOQ