C10 / 6.12.

2012

Modeling and Analysis

Management support system modeling Management support system modeling Static and Dynamic Models Static and Dynamic Models Certainty, Uncertainty, and Risk Certainty, Uncertainty, and Risk Management Support Systems Modeling with Spreadsheets Management Support Systems Modeling with Spreadsheets Decision Tables Decision Tables Decision Tree Decision Tree The structure of Mathematical Models for Decision Support The structure of Mathematical Models for Decision Support Mathematical Programming Optimization Mathematical Programming Optimization Multiple Goals, Sensitivity Analysis, What-If Analysis, and Multiple Goals, Sensitivity Analysis, What-If Analysis, and Goal Seeking Goal Seeking 10. Problem-Solving Search Methods 10. Problem-Solving Search Methods 11. Simulation Applications ) 11. Simulation Applications (()
1/47

1. 1. 2. 2. 3. 3. 4. 4. 5. 5. 6. 6. 7. 7. 8. 8. 9. 9.

Modeling and Analysis Modeling and Analysis

Learning objectives: Learning objectives:
Understand the basic concepts of management support system MSS) modeling Understand the basic concepts of management support system ((MSS)modeling MSS Describe how MSS models interact with data and user Describe how MSS models interact with data and user Understand some different well-known model classes Understand some different well-known model classes Understand how to structure decision making with few alternatives Understand how to structure decision making with aafew alternatives Describe how spreadsheets can be used for MSS modeling and solution Describe how spreadsheets can be used for MSS modeling and solution Explain the basic concepts of optimization, simulation, and heuristics, and when Explain the basic concepts of optimization, simulation, and heuristics, and when to use them to use them Describe how to structure linear programming model Describe how to structure aalinear programming model Understand how search methods are used to solve MSS models Understand how search methods are used to solve MSS models Explain the differences among algorithms, blind search, and heuristics Explain the differences among algorithms, blind search, and heuristics Describe how to handle multiple goals Describe how to handle multiple goals Explain what is meant by sensitivity analysis, what-if analysis, and goal seeking Explain what is meant by sensitivity analysis, what-if analysis, and goal seeking Describe the key issues of model management Describe the key issues of model management
2/47

Management support system modeling Management support system modeling

Modeling is key element in most DSS and necessity in model-based DSS DSS. Modeling is aakey element in most DSS and aanecessity in aamodel-based DSS. There are many classes of models, and there are often many specialized techniques There are many classes of models, and there are often many specialized techniques for solving each one. Simulation is common modeling approach, but there are for solving each one. Simulation is aa common modeling approach, but there are several others: several others: A general model based on an algorithm (for example to make transportationmodel, A general model, based on an algorithm (for example to make transportationcost estimates). This model is programmed directly into theDSS DSS. cost estimates). This model is programmed directly into the DSS. A demand-forecasting model (statistics based). A demand-forecasting model (statistics based). A distribution center locationmodel This model uses aggregated data (a special model. A distribution center location model. This model uses aggregated data (a special modeling technique) and is solved with standard linear/integer optimization modeling technique) and is solved with aa standard linear/integer optimization package. package. A transportation model (i.e. specialization of linear programming model) to A transportation model (i.e. aaspecialization of aalinear programming model) to determine the best shipping option from sources to distribution centers (fed to it determine the best shipping option from sources to distribution centers (fed to it from the previous model) and hence to customers. This model is solved using from the previous model) and hence to customers. This model is solved using commercial software and is loosely integrated with the distribution location commercial software and is loosely integrated with the distribution location model. The DSS must interface with commercial software and integrate the model. The DSS must interface with commercial software and integrate the model. model. A financial and risk simulation model that takes into consideration some A financial and risk simulation model that takes into consideration some qualitative factors that require important human judgment. qualitative factors that require important human judgment. A geographical information system (GIS; effectively graphical model of the A geographical information system (GIS; effectively aa graphical model of the data) for user interface. data) for aauser interface.
3/47

Some major modeling issues Some major modeling issues

1. Identification of the problem and Environmental Analysis 1. Identification of the problem and Environmental Analysis
An important aspect of the environmental analysis is environmental scanning and An important aspect of the environmental analysis is environmental scanning and analysis, which is the monitoring, scanning, and interpretation of collected analysis, which is the monitoring, scanning, and interpretation of collected information. No decision is made in vacuum. It is important to analyze the scope information. No decision is made in aavacuum. It is important to analyze the scope of the domain and the forces and dynamics of the environment. A decision maker of the domain and the forces and dynamics of the environment. A decision maker needs to identify the organizational culture and the corporate decision-making needs to identify the organizational culture and the corporate decision-making process. It is entirely possible that environmental factors have created the current process. It is entirely possible that environmental factors have created the current problem. The problem must be understood, and everyone involved should share problem. The problem must be understood, and everyone involved should share the same frame of understanding because the problem will ultimately be the same frame of understanding because the problem will ultimately be represented by the model in one form or another. Otherwise, the model will not represented by the model in one form or another. Otherwise, the model will not help the decision maker. help the decision maker.

2. Variable identification 2. Variable identification

Identification of models variables is critical, as are relationships of the Identification of aa models variables is critical, as are relationships of the variables. Influence diagrams, which are graphical models of mathematical models, variables. Influence diagrams, which are graphical models of mathematical models, can facilitate the identification process. A more general form of an influence can facilitate the identification process. A more general form of an influence diagram, cognitive map, can help decision maker develop better understanding diagram, aacognitive map, can help aadecision maker develop aabetter understanding of problem, especially of variables and their interactions. of aaproblem, especially of variables and their interactions.
4/47

 Some major modeling issues Some major modeling issues

3. Forecasting Predictive Analytics 3. Forecasting Predictive Analytics

Forecasting is predicting the future. This form of predictive analytics is esential Forecasting is predicting the future. This form of predictive analytics is esential for construction and manipulating models because when decision is implemented, for construction and manipulating models because when aadecision is implemented, the results usually occur in the future. Whereas DSS are typically designed to the results usually occur in the future. Whereas DSS are typically designed to determine what will be, traditional MIS report what is or what was. There is no determine what will be, traditional MIS report what is or what was. There is no point in running what-if (sensitivity) analysis on the past because decisions made point in running aawhat-if (sensitivity) analysis on the past because decisions made then have no impact in the future. Forecasting is getting easier as software vendors then have no impact in the future. Forecasting is getting easier as software vendors automate many of the complications of developing such models. Forecasting automate many of the complications of developing such models. Forecasting system that incorporates its predictive analytics technology, ideally for retailers. system that incorporates its predictive analytics technology, ideally for retailers. This software iss more automated than most other forecasting packages. This software iss more automated than most other forecasting packages.

4. Multiple Models 4. Multiple Models

A DSS can include several models which represents different part of the A DSS can include several models which represents aa different part of the decision-making problem (for example P&G supply-chain DSS includes location decision-making problem (for example P&G supply-chain DSS includes aalocation model to locate distribution center, product-strategy model, demand-forecasting model to locate distribution center, aaproduct-strategy model, aademand-forecasting model, cost generation model, financial- and risk-simulation model, and even model, aacost generation model, aafinancial- and risk-simulation model, and even aa GIS model). GIS model).
5/47

 Some major modeling issues Some major modeling issues

5. Model Categories 5. Model Categories

There are seven groups of DSS models. Each category can be applied to either There are seven groups of DSS models. Each category can be applied to either static or dynamic model, which can be constructed under assumed aa static or aa dynamic model, which can be constructed under assumed environments of certainty, uncertainty, or risk. To expedite model construction, environments of certainty, uncertainty, or risk. To expedite model construction, we can use special decision analysis systems that have modeling languages and we can use special decision analysis systems that have modeling languages and capabilities embedded in them. These includes spreadsheets, data mining capabilities embedded in them. These includes spreadsheets, data mining systems, OLAP systems, and even fourth-generation languages (formerly systems, OLAP systems, and even fourth-generation languages (formerly financial planning languages). financial planning languages).

6/47

 5. Model Categories 5. Model Categories Category

Optimization of problems with few alternatives Optimization via algorithm Optimization via an analytic formula Simulation Heuristics Predictive models Other models

Some major modeling issues Some major modeling issues Representative Techniques
Decision tables, decision tree Linear programming models, network models Some inventory models

Process and Objectives

Find the best solution from a small number of alternatives Find the best solution from a large number of alternatives, using a stepby-step improvement process Find the best solution in one step, using a formula Finding a good enough solution or the best among the alternatives checked, using experimentation Find a good enough solution, using rules Predict the future for a given scenario Solve a what-if case, using a formula

Several types of simulation Heuristic programming, expert systems Forecasting models, Markov analysis Financial modeling, waiting lines
7/47

 Some major modeling issues Some major modeling issues

6. Model Management 6. Model Management

Models, like data, must be managed to maintain their integrity and thus their Models, like data, must be managed to maintain their integrity and thus their applicability. Such management is done with the aid of the model base applicability. Such management is done with the aid of the model base management systems MBMS), which are analogous to database management management systems ((MBMS), which are analogous to database management MBMS systems DBMS) systems ((DBMS). . DBMS

7. Knowledge-Based Modeling 7. Knowledge-Based Modeling

DSS uses mostly quantitative models, whereas expert systems use qualitative, DSS uses mostly quantitative models, whereas expert systems use qualitative, knowledge-based models in their applications. Some knowledge is necessary to knowledge-based models in their applications. Some knowledge is necessary to construct solvable (and therefore usable) models. We defer the description of construct solvable (and therefore usable) models. We defer the description of knowledge-based models until later chapters. knowledge-based models until later chapters.

8/47

8. Current Trends in Modeling 8. Current Trends in Modeling

Some major modeling issues Some major modeling issues

One recent trend in modeling involves the development of model libraries and One recent trend in modeling involves the development of model libraries and solution technique libraries. Some of these codes can be run directly on the solution technique libraries. Some of these codes can be run directly on the owners web server for free, and others can be downloaded and run on an owners web server for free, and others can be downloaded and run on an individuals PC, Unix machine, or server. The availability of these codes means individuals PC, Unix machine, or server. The availability of these codes means that powerful optimization and simulation package are available to decision makers that powerful optimization and simulation package are available to decision makers who may have only experienced these tools from the perspective of classroom who may have only experienced these tools from the perspective of classroom problems. problems. There is clear trend toward developing and using Web tools and software to There is aa clear trend toward developing and using Web tools and software to access and even run software to perform modeling, optimization, simulation, and so access and even run software to perform modeling, optimization, simulation, and so on. This has, in many ways, simplified the application of many models to realon. This has, in many ways, simplified the application of many models to realworld problems. Another trend, involves the lack of understanding of what models world problems. Another trend, involves the lack of understanding of what models and their solutions can do in the real world. and their solutions can do in the real world. There is continuing trend toward making MSS models completely transparent There is aa continuing trend toward making MSS models completely transparent to decision maker. For example, multidimensional analysis (modeling) involves to decision maker. For example, multidimensional analysis (modeling) involves data analysis in several dimensions. data analysis in several dimensions. There is also trend to model model to help in its analysis. An influence There is also aa trend to model aa model to help in its analysis. An influence diagram is graphical representation of model; that is, it is model of model. diagram is aagraphical representation of aamodel; that is, it is aamodel of aamodel. Some influence diagram software packages are capable of generating and solving Some influence diagram software packages are capable of generating and solving the resultant model. the resultant model.
9/47

Static and Dynamic Models Static and Dynamic Models

Dss models can be classified as static ordynamic dynamic: Dss models can be classified as static or dynamic: Static Analysis: Static Analysis: A static model takes single snapshot of situation. During this snapshot, A static model takes aa single snapshot of aa situation. During this snapshot, everything occurs in single interval. For example, decision about whether to everything occurs in aa single interval. For example, aa decision about whether to make or buy product is static in nature. make or buy aaproduct is static in nature. Dynamic Analysis: Dynamic Analysis: Dynamic models represent scenarios that change over time. For example, 5Dynamic models represent scenarios that change over time. For example, aa5years profit-and-loss projections in which the input data (costs, prices, and years profit-and-loss projections in which the input data (costs, prices, and quantities) change from year to year. Dynamic models are time dependent dependent. quantities) change from year to year. Dynamic models are time dependent. Dynamic models use, represent, or generate trends and patters over time, show Dynamic models use, represent, or generate trends and patters over time, show averages per periods, moving averages, and comparative analyses. When static averages per periods, moving averages, and comparative analyses. When aastatic model is constructed to describe given situation, it can be expanded to represent model is constructed to describe aagiven situation, it can be expanded to represent the dynamic nature of the problem. the dynamic nature of the problem.
10/47

Certainty, Uncertainty, and Risk Certainty, Uncertainty, and Risk

Decision making process involves evaluating and comparing alternatives. Decision making process involves evaluating and comparing alternatives. During the process, it is necessary to predict the future outcome of each proposed During the process, it is necessary to predict the future outcome of each proposed alternative. Decision situation are often classified on the basis of what the decision alternative. Decision situation are often classified on the basis of what the decision maker knows or believes about the forecasted result. We customarily classify this maker knows or believes about the forecasted result. We customarily classify this knowledge into three categories, ranging from complete knowledge to total knowledge into three categories, ranging from complete knowledge to total ignorance: ignorance: Certainty Certainty Risk Risk Uncertainty Uncertainty When we are develop models, any of these conditions can occur, and different When we are develop models, any of these conditions can occur, and different kind of models are appropriate for each case. kind of models are appropriate for each case. Increasing knowledge Complete knowledge certainly Ignorance, total uncertainly
11/47

Risk Decreasing knowledge

 Certainty, Uncertainty, and Risk Certainty, Uncertainty, and Risk

Decision making under certainty Decision making under certainty

In decision making under certainty, it is assumed that complete knowledge is In decision making under certainty, it is assumed that complete knowledge is available the decision maker knows exactly what the outcome of each course of available the decision maker knows exactly what the outcome of each course of action will be (a deterministic environment). action will be (a deterministic environment). The decision maker is viewed as perfect predictor on the future. The decision maker is viewed as aaperfect predictor on the future. Certainly models are relatively easy to develop and solve, and they can yield Certainly models are relatively easy to develop and solve, and they can yield optimal solution. optimal solution.

Decision making under uncertainty Decision making under uncertainty

In decision making under uncertainty, the decision maker considers situations in In decision making under uncertainty, the decision maker considers situations in which several outcomes are possible for each course of action. The decision maker which several outcomes are possible for each course of action. The decision maker does not know, or cannot estimate, the probability of occurrence of the possible does not know, or cannot estimate, the probability of occurrence of the possible outcomes. outcomes. The problem, in this case, is more difficult because there is insufficient The problem, in this case, is more difficult because there is insufficient information modeling involves assessment of the decision makers attitude information modeling involves assessment of the decision makers attitude toward risk. toward risk.
12/47

 Certainty, Uncertainty, and Risk Certainty, Uncertainty, and Risk

Decision making under risk Decision making under risk

In decision made under risk (a probabilistic or stochastic decision-making In decision made under risk (a probabilistic or stochastic decision-making situation), is one in which the decision maker must consider several possible situation), is one in which the decision maker must consider several possible outcomes for each alternative, each with given probability of occurrence. outcomes for each alternative, each with aagiven probability of occurrence. The decision maker can asses the degree of risk associated with each The decision maker can asses the degree of risk associated with each alternative (calculated risk). alternative (calculated risk). Risk analysis is decision-making method that analyzes the risk (based on Risk analysis is aa decision-making method that analyzes the risk (based on assumed known probabilities) associated with different alternatives. assumed known probabilities) associated with different alternatives. Risk analysis can be performed by calculating the expected value of each Risk analysis can be performed by calculating the expected value of each alternative and selecting the one with the best expectedvalue value. alternative and selecting the one with the best expected value.

13/47

Management Support Systems Modeling with Spreadsheets Management Support Systems Modeling with Spreadsheets
Models can be developed and implemented in variety of programming Models can be developed and implemented in aa variety of programming languages and systems: from third-, fourth-, and fifth-generation to computer aided languages and systems: from third-, fourth-, and fifth-generation to computer aided software engineering (CASE) systems and other systems that automatically software engineering (CASE) systems and other systems that automatically generate usable software. We focus primarily on spreadsheets, model languages, generate usable software. We focus primarily on spreadsheets, model languages, and transparent data analysis tools the most popular end-user modeling tool. and transparent data analysis tools the most popular end-user modeling tool. Spreadsheets include extensive statistical, forecasting, and other modeling and Spreadsheets include extensive statistical, forecasting, and other modeling and database management capabilities, functions, and routines Solver, Whats Best, database management capabilities, functions, and routines ((Solver, Whats Best, Braincel, NeuralTools, Evolver,@Risk @Risk). Braincel, NeuralTools, Evolver, @Risk).
Loan Calculation Model Loan Amount Interest Rate Number of Years Number of Month Interest Rate/Month Monthly Loan Payment $150,000 8.00% 30 360 0.67% -$1,100.65
Month 0 1 2 3 4 5 $1,100.65 $1,100.65 $1,100.65 $1,100.65 $1,100.65 $100.00 $100.00 $100.00 $100.00 $100.00 $1,200.65 $1,200.65 $1,200.65 $1,200.65 $1,200.65 Normal Paym.

Dynamic Model
$100.00 Prepay Amount Total Payment Principale Owed $150,000 $149,799.35 $149,597.37 $149,394.04 $149,189.35 $148,983.30

14/47

5. Decision Tables 5. Decision Tables

Decision tables organize information and knowledge in systematic, tabular Decision tables organize information and knowledge in aa systematic, tabular manner to prepare it for analysis. manner to prepare it for analysis. Example. Example. An investment company is considering investing in one of three alternatives: An investment company is considering investing in one of three alternatives: bonds, stocks, or certificates of deposit (CDs). The company is interested in one bonds, stocks, or certificates of deposit (CDs). The company is interested in one goal: maximizing the yield on the investment after one year. If it were interested in goal: maximizing the yield on the investment after one year. If it were interested in other goals, such as safety or liquidity, the problem would be classified as one of other goals, such as safety or liquidity, the problem would be classified as one of multi-criteria decision analysis. multi-criteria decision analysis. The yield depends on the state of the economy sometime in the future (called the The yield depends on the state of the economy sometime in the future (called the state of nature), which can be in solid growth, stagnation, or inflation. Experts state of nature), which can be in solid growth, stagnation, or inflation. Experts estimated the following annual yields: estimated the following annual yields: If there is solid growth in the economy, bonds will yield 12%, stocks 15%, and If there is solid growth in the economy, bonds will yield 12%, stocks 15%, and time deposits 6.5%. time deposits 6.5%. If stagnation prevails, bonds will yield 6%, stocks 3%, and time deposits 6.5%. If stagnation prevails, bonds will yield 6%, stocks 3%, and time deposits 6.5%. If inflation prevails, bonds will yield 3%, stocks will bring loss of 2%, and If inflation prevails, bonds will yield 3%, stocks will bring aa loss of 2%, and time deposits will yield 6.5%. time deposits will yield 6.5%.
15/47

 5. Decision Tables 5. Decision Tables The problem is to select the one best investment alternative. These are assumed The problem is to select the one best investment alternative. These are assumed to be discrete alternatives. to be discrete alternatives. The investment decision-making problem can be viewed as two-person game. The investment decision-making problem can be viewed as aatwo-person game. the investor makes choice (a move), and then state of nature occurs (makes the investor makes aa choice (a move), and then aa state of nature occurs (makes aa move). The next table shows the payoff of mathematical model. The table move). The next table shows the payoff of aa mathematical model. The table includes decision variables (the alternatives), uncontrollable variables, and result includes decision variables (the alternatives), uncontrollable variables, and result variables (the project yields). variables (the project yields). Investment Problem Decision Table Model State of Nature (Un controllable Variables)
-

Alternative Bonds (Obligat.) Stocks (Actiuni) CDs (Cert. Dep.)

Solid Growth 12.0% 15.0% 6.5%

Stagnation 6.0% 3.0% 6.5%

Inflation 3.0% -2.0% 6.5%

16/47

 5. Decision Tables 5. Decision Tables

Treating Uncertainty Treating Uncertainty

There are several methods of handling uncertainty: There are several methods of handling uncertainty: The optimistic approach assumes that the best possible outcome of each The optimistic approach assumes that the best possible outcome of each alternative will occur and then selects the best of the best (Stocks). alternative will occur and then selects the best of the best (Stocks). The pessimistic approach assumes that the worst possible outcome for The pessimistic approach assumes that the worst possible outcome for each alternative will occur and selects the best of these (CDs). each alternative will occur and selects the best of these (CDs). Another approach simply assumes that all states of nature are equally Another approach simply assumes that all states of nature are equally possible. possible. possible The problem can be treated under assumed certainty or risk. The problem can be treated under assumed certainty or risk.

Alternative Bonds (Obligat.) Stocks (Actiuni) CDs (Cert. Dep.)

Solid Growth 12.0% 15.0% 6.5%

Stagnation 6.0% 3.0% 6.5%

Inflation 3.0% -2.0% 6.5%

17/47

Treating Risk Treating Risk

5. Decision Tables 5. Decision Tables

The most common method for solving this risk analysis problem is to select the The most common method for solving this risk analysis problem is to select the alternative with the greatest expected value. alternative with the greatest expected value. Assume that the experts estimate the chance of solid growth at 50%, the chance of Assume that the experts estimate the chance of solid growth at 50%, the chance of stagnation at 30%, and the chance of inflation at 20%. The decision table is then stagnation at 30%, and the chance of inflation at 20%. The decision table is then rewritten with the known probabilities (see next table). An expected value is rewritten with the known probabilities (see next table). An expected value is computed by multiplying the results (outcomes) by their respective probabilities computed by multiplying the results (outcomes) by their respective probabilities and adding them. For example, investing in bonds yields an expected return of and adding them. For example, investing in bonds yields an expected return of 12(0,5)+6(0.3)+3(0.2)=8.4%. (This approach can be dangerous strategy!) 12(0,5)+6(0.3)+3(0.2)=8.4%. (This approach can be aadangerous strategy!) Decision Under Risk and its Solution Solid Growth Alternative Bonds Stocks CDs 50.0% 12.0% 15.0% 6.5% Stagnation 30.0% 6.0% 3.0% 6.5% Inflation 20.0% 3.0% -2.0% 6.5% Expected Value 8.4% 8.0% 6.5%
18/47

6. Decision Trees 6. Decision Trees

An alternative representation of the decision table is decision tree (see An alternative representation of the decision table is aa decision tree (see Mind Tools). Mind Tools). A decision tree shows the relationships of the problem graphically and can A decision tree shows the relationships of the problem graphically and can handle complex situations in compact form. handle complex situations in aacompact form. A decision tree can be cumbersome (ancombrant) if there are many A decision tree can be cumbersome (ancombrant) if there are many alternatives or states of nature. alternatives or states of nature. Tree Age Pro and Precision Tree include powerful, intuitive, and Tree Age Pro and Precision Tree include powerful, intuitive, and sophisticated decision tree analysis systems. sophisticated decision tree analysis systems. These vendors also provide excellent examples of decision tree used in These vendors also provide excellent examples of decision tree used in practice (psychwww.com/mtsite/dectree.html ). practice (psychwww.com/mtsite/dectree.html ??).
19/47

 6. Decision Trees 6. Decision Trees A simplified investment case of multiple goals (a decision situation in which A simplified investment case of multiple goals (a decision situation in which alternatives are evaluated with several, sometimes conflicting, goal) is shown in alternatives are evaluated with several, sometimes conflicting, goal) is shown in next table. next table.
Multiple Goals Alternative Bonds Stocks CDs Yield
8.4% 8.0% 6.5%

Safety
High Low Very High

Liquidity
High High High

The three goals criteria) are yield, safety, liquidity This situation is under liquidity. The three goals ((criteria) are yield, safety, liquidity. This situation is under criteria assumed certainty; that is, only one possible consequence is projected for each assumed certainty; that is, only one possible consequence is projected for each alternative; the more complex cases of risk or uncertainty could be considered. alternative; the more complex cases of risk or uncertainty could be considered. some of the results are qualitative (e.g. low, high) rather than numeric. some of the results are qualitative (e.g. low, high) rather than numeric.
20/47

7. The structure of Mathematical Models for Decision Support 7. The structure of Mathematical Models for Decision Support
The Components of decision Support Mathematical Models The Components of decision Support Mathematical Models
All models are made up of four components: result, decision, uncontrollable, All models are made up of four components: result, decision, uncontrollable, and intermediate result variables: and intermediate result variables:
Uncontrollable Variables Uncontrollable Variables Decision Variables Decision Variables Mathematical Relationships Mathematical Relationships Intermediate Variables Intermediate Variables Result Variables Result Variables

Mathematical Relationships link these components together. The modeling Mathematical Relationships link these components together. The modeling process involves identifying the variables and relationships among them. Solving process involves identifying the variables and relationships among them. Solving aa model determines the values of these and the result variables. model determines the values of these and the result variables.
21/47

 7. The structure of Mathematical Models for Decision Support 7. The structure of Mathematical Models for Decision Support
The Components of decision Support Mathematical Models The Components of decision Support Mathematical Models

Result (Outcome) Variables Result (Outcome) Variables Result (outcome) variables reflect the level of effectiveness of system; that is, Result (outcome) variables reflect the level of effectiveness of aasystem; that is, they indicate how well the system performs or attains its goal. they indicate how well the system performs or attains its goal. Result variables (outputs) are considered dependent variables. Result variables (outputs) are considered dependent variables. Intermediate result variables are sometimes used in modeling to identify Intermediate result variables are sometimes used in modeling to identify intermediate outcomes. intermediate outcomes. In the case of dependent variable, another event must occur first before the In the case of aa dependent variable, another event must occur first before the event described by the variable can occur. event described by the variable can occur. Result variables depend on the occurrence of the decision variables and the Result variables depend on the occurrence of the decision variables and the uncontrollable variables. uncontrollable variables.
22/47

 7. The structure of Mathematical Models for Decision Support 7. The structure of Mathematical Models for Decision Support
Result (Outcome) Variables Result (Outcome) Variables The Components of decision Support Mathematical Models The Components of decision Support Mathematical Models

Examples of the Components of Models

Area Financial investment Marketing Manufacturing Accounting Transportation Services

Decision Variables

Result Variables Total profit, risk Rate of return of investment Earnings per share Liquidity level Market share Customer satisfaction Total cost Quality level Employee satisfaction Data processing cost Error rate Total transport cost Payment float time Customer satisfaction

Uncontrollable Variables and Parameters Inflation rate Prime rate Competition

Investment alternatives and amounts

Advertising budget Where to advertise What and how much to produce Inventory levels Compensation programs Use of computers Audit schedule Shipments schedules Use smart cards Staffing levels

Customers income Competitors actions Machine capacity Technology Materials prices Computer technology Tax rates Legal requirements Delivery distance Regulations Demand for services

23/47

 7. The structure of Mathematical Models for Decision Support 7. The structure of Mathematical Models for Decision Support
The Components of decision Support Mathematical Models The Components of decision Support Mathematical Models

Decision Variables Decision Variables Decision variables describe alternative courses of action. Decision variables describe alternative courses of action. The decision maker controls the decision variables. The decision maker controls the decision variables. For example, for an investment problem, the amount to invest in bonds is For example, for an investment problem, the amount to invest in bonds is aa decision variable. decision variable. In scheduling problem, the decision variables are people, times, and schedule. In aascheduling problem, the decision variables are people, times, and schedule. Other examples are listed in the next table. Other examples are listed in the next table.

24/47

 7. The structure of Mathematical Models for Decision Support 7. The structure of Mathematical Models for Decision Support
Decision Variables Decision Variables The Components of decision Support Mathematical Models The Components of decision Support Mathematical Models Examples of the Components of Models

Area Financial investment Marketing Manufacturing Accounting Transportation Services

Decision Variables Investment alternatives and amounts Advertising budget Where to advertise What and how much to produce Inventory levels Compensation programs Use of computers Audit schedule Shipments schedules Use smart cards Staffing levels

Result Variables Total profit, risk Rate of return of investment Earnings per share Liquidity level Market share Customer satisfaction Total cost Quality level Employee satisfaction Data processing cost Error rate Total transport cost Payment float time Customer satisfaction

Uncontrollable Variables and Parameters Inflation rate Prime rate Competition Customers income Competitors actions Machine capacity Technology Materials prices Computer technology Tax rates Legal requirements Delivery distance Regulations Demand for services

25/47

 7. The structure of Mathematical Models for Decision Support 7. The structure of Mathematical Models for Decision Support
The Components of decision Support Mathematical Models The Components of decision Support Mathematical Models

Uncontrollable Variables, or Parameters Uncontrollable Variables, or Parameters In any decision-making situation, there are factors that affect the result In any decision-making situation, there are factors that affect the result variables but are not under the control of the decision maker. Either these factors variables but are not under the control of the decision maker. Either these factors can be fixed, in which case they are called uncontrollable variables, or parameters, can be fixed, in which case they are called uncontrollable variables, or parameters, or they can vary, in which case they are called variables. or they can vary, in which case they are called variables. Examples of factors are the prime interest rate, citys building code, tax Examples of factors are the prime interest rate, aa citys building code, tax regulations, and utilities costs. Most of these factors are uncontrollable because regulations, and utilities costs. Most of these factors are uncontrollable because they are in and determined by elements of the system environment in which the they are in and determined by elements of the system environment in which the decision maker works. Some of these variables limit the decision maker and decision maker works. Some of these variables limit the decision maker and therefore form what are called the constrains of the problem. therefore form what are called the constrains of the problem.

26/47

 7. The structure of Mathematical Models for Decision Support 7. The structure of Mathematical Models for Decision Support
Decision Variables Decision Variables The Components of decision Support Mathematical Models The Components of decision Support Mathematical Models Examples of the Components of Models

Area Financial investment Marketing Manufacturing Accounting Transportation Services

Decision Variables

Result Variables Total profit, risk Rate of return of investment Earnings per share Liquidity level Market share Customer satisfaction Total cost Quality level Employee satisfaction

Uncontrollable Variables and Parameters Inflation rate Prime rate Competition Customers income Competitors actions Machine capacity Technology Materials prices Computer technology Tax rates Legal requirements Delivery distance Regulations Demand for services
27/47

Investment alternatives and amounts Advertising budget Where to advertise What and how much to produce Inventory levels Compensation programs

Use of computers Audit schedule

Data processing cost Error rate

Shipments schedules Use smart cards Staffing levels

Total transport cost Payment float time Customer satisfaction

 7. The structure of Mathematical Models for Decision Support 7. The structure of Mathematical Models for Decision Support
The Components of decision Support Mathematical Models The Components of decision Support Mathematical Models

Intermediate Result Variables Intermediate Result Variables Intermediate result variable reflect intermediate outcomes in mathematical Intermediate result variable reflect intermediate outcomes in mathematical models. models. For example, in determining machine scheduling, spoilage is an For example, in determining machine scheduling, spoilage is an intermediate result variable, and total profit is the result variable (i.e., spoilage intermediate result variable, and total profit is the result variable (i.e., spoilage is one determinant of total profit). is one determinant of total profit). Another example is employee salaries. This constitutes decision variable Another example is employee salaries. This constitutes aa decision variable for management. It determines employee satisfaction (i.e., intermediate for management. It determines employee satisfaction (i.e., intermediate outcome), which in turn determines the productivity level (i.e., final result). outcome), which in turn determines the productivity level (i.e., final result).

28/47

 7. The structure of Mathematical Models for Decision Support 7. The structure of Mathematical Models for Decision Support
The Components of decision Support Mathematical Models The Components of decision Support Mathematical Models

The structure of MSS Mathematical Models The structure of MSS Mathematical Models The component of quantitative model are linked together by mathematical The component of aa quantitative model are linked together by mathematical (algebraic) expressions equations or inequalities. (algebraic) expressions equations or inequalities. A very simple financial model is A very simple financial model is where P = profit, R = revenue, and C = cost. This equation describes the where P = profit, R = revenue, and C = cost. This equation describes the relationship among the variables. relationship among the variables. Another well-known financial model is the simple present-value cash flow Another well-known financial model is the simple present-value cash flow model, where P = present value, F = future single payment in dollars, = interest model, where P = present value, F = aafuture single payment in dollars, ii= interest rate (percent-age), and = number of years. With this model it is possible to rate (percent-age), and nn = number of years. With this model it is possible to determine the present value of payment of $100,000 to be made five years from determine the present value of aapayment of $100,000 to be made five years from today, at 10% (0.1) interest rate, as follows: today, at aa10% (0.1) interest rate, as follows: P = 100,000 (1+0.1)5 = $62,092 P = 100,000 //(1+0.1)5 = $62,092
29/47

P=RC P=RC

8. Mathematical Programming Optimization 8. Mathematical Programming Optimization

Linear programming LP) is the best-know technique in family of optimization Linear programming ((LP)is the best-know technique in family of optimization LP tools called mathematical programming In LP all relationships among the programming. LP, tools called mathematical programming. In LP, all relationships among the variables are linear. variables are linear.

Mathematical Programming Mathematical Programming

Mathematical programming is family of tools designed to help solve Mathematical programming is aa family of tools designed to help solve managerial problems in which the decision maker must allocate scarce resources managerial problems in which the decision maker must allocate scarce resources among competing activities to optimize measurable goal. among competing activities to optimize aameasurable goal.

30/47

 8. Mathematical Programming Optimization 8. Mathematical Programming Optimization

Linear Programming Linear Programming

Every LP problem is composed of: Every LP problem is composed of: decision variable whose values are unknown and are searched for, decision variable --whose values are unknown and are searched for, an objective function linear mathematical function that relates the decision an objective function --aalinear mathematical function that relates the decision variables to the goal, measures goal attainment, and is to be optimized, variables to the goal, measures goal attainment, and is to be optimized, objective function coefficients -unit profit or cost coefficients indicating the objective function coefficients -unit profit or cost coefficients indicating the contribution to the objective of one unit of decision variable, contribution to the objective of one unit of aadecision variable, constraints expressed in the form of linear inequalities or equalities that constraints -- expressed in the form of linear inequalities or equalities that limit resources or requirements; these relate the variables through linear limit resources or requirements; these relate the variables through linear relationships, relationships, capacities which describe the upper and sometimes lower limits on the capacities -- which describe the upper and sometimes lower limits on the constraints and variables, and constraints and variables, and input/output (technology) coefficients which indicate resource utilization for input/output (technology) coefficients --which indicate resource utilization for decision variable. aadecision variable.
31/47

 8. Mathematical Programming Optimization 8. Mathematical Programming Optimization

Linear Programming Linear Programming

An example of modeling in LP: An example of modeling in LP: The decision variables The decision variables :: X= units of CC-7 X11=units of CC-7 X= units of CC-8 X 2=units of CC-8
2

The result variable The result variable ::

Total profit = Z Total profit = Z The objective is to maximize The objective is to maximize :: Z = 8,000X+12,000X Z = 8,000X11+12,000X2 2 The uncontrollable variables ((constraints):: The uncontrollable variables constraints) constraints Labor constrain: 300X+ 500X 200,000 Labor constrain: 300X 1+ 500X 2 200,000 Budget constrain: 10,000X+15,000X 8,000,000 Budget constrain: 10,000X11+15,000X2 8,000,000 2 Marketing requirement for CC-7: X 100 Marketing requirement for CC-7: X 1100 Marketing requirement for CC-8: X 200 Marketing requirement for CC-8: X22200
32/47
1 1 2

 8. Mathematical Programming Optimization 8. Mathematical Programming Optimization

Linear Programming Linear Programming

The Mathematical Model is The Mathematical Model is::

Decision Decision
Variables Variables X= units of CC-7 X11=units of CC-7 X= units of CC-8 X 2=units of CC-8
2

Mathematical Mathematical Relationships Relationships Maximize Z (profit) Maximize Z (profit) subject to constraints subject to constraints Constraints Constraints (Uncontrollable) (Uncontrollable)
300X+500X 200,000 300X11+500X2 200,000 2 100X+1,5000X 8,000,000 100X11+1,5000X2 8,000,000 2 X 100 X11100 X 200 X 2200
2

Result Result Variables Variables

Total profit = Z Total profit = Z Z = 8,000X+12,000X Z = 8,000X 1+12,000X2
1 2

33/47

 8. Mathematical Programming Optimization 8. Mathematical Programming Optimization

Linear Programming Linear Programming

The Excel Solver Solution is The Excel Solver Solution is::

X1 Decision Variables: Total Profit: Labor: Budget: X1 lower: X2 lower: 333.33 8.00 0.30 10.00 1.00 0.00

X2 200.00 12.00 0.50 15.00 0.00 1.00 5,066.67 200.00 6,333.33 333.33 200.00 200.00 8,000.00 100.00 200.00

34/47

 8. Mathematical Programming Optimization 8. Mathematical Programming Optimization

The Excel Solver Answer Report is The Excel Solver Answer Report is::
Microsoft Excel 11.0 Answer Report Worksheet: [PL.xls]PL Report Created: 12/13/2009 7:57:12 PM

Linear Programming Linear Programming

Target Cell (Max) Cell $E$7 Total Profit Name Original Value 5,066.67 Final Value 5,066.67

Adjustable Cells Cell $C$6 $D$6 Name Decision Variables: X1 Decision Variables: X2 Original Value 333.33 200.00 Final Value 333.33 200.00

Constraints Cell $E$8 $E$9 $E$10 $E$11 Labor: Budget: X1 lower: X2 lower: Name Cell Value 200.00 6,333.33 333.33 200.00 Formula $E$8<=$F$8 $E$9<=$F$9 $E$10>=$F$10 $E$11>=$F$11 Status Binding Not Binding Not Binding Binding Slack 0 1666.666667 233.33 0.00

35/47

 8. Mathematical Programming Optimization 8. Mathematical Programming Optimization

Linear Programming Linear Programming

The Excel Solver Sensitivity Report is The Excel Solver Sensitivity Report is::
Microsoft Excel 11.0 Sensitivity Report Worksheet: [PL.xls]PL Report Created: 12/13/2009 7:57:12 PM

Adjustable Cells Final Cell $C$6 $D$6 Decision Variables: X1 Decision Variables: X2 Name Value 333.33 200.00 Reduced Gradient 0.00 0.00

Constraints Final Cell $E$8 $E$9 $E$10 $E$11 Labor: Budget: X1 lower: X2 lower: Name Value 200.00 6,333.33 333.33 200.00 Lagrange Multiplier 26.67 0.00 0.00 -1.33

36/47

 8. Mathematical Programming Optimization 8. Mathematical Programming Optimization

Linear Programming Linear Programming

The Excel Solver Limits Report is The Excel Solver Limits Report is::
Microsoft Excel 11.0 Limits Report Worksheet: [PL.xls]Limits Report 2 Report Created: 12/13/2009 7:57:12 PM

Target Cell $E$7 Total Profit Name Value 5,066.67

Adjustable Cell $C$6 $D$6 Name Decision Variables: X1 Decision Variables: X2 Value 333.33 200.00

Lower Limit 100.00 200.00

Target Result 3,200.00 5,066.67

Upper Limit 333.33 200.00

Target Result 5,066.67 5,066.67

37/47

 8. Mathematical Programming Optimization 8. Mathematical Programming Optimization

Linear Programming Linear Programming

The most common optimization models can be solved by variety of The most common optimization models can be solved by aa variety of mathematical programmingmethods including the following: methods, mathematical programming methods, including the following: Assignment (best matching of objects) Assignment (best matching of objects) Dynamic programming Dynamic programming Goal programming Goal programming Investment (maximizing rate of return) Investment (maximizing rate of return) Linear and integer programming Linear and integer programming Network models for planning and scheduling Network models for planning and scheduling Nonlinear programming Nonlinear programming Replacement (capital budgeting) Replacement (capital budgeting) Simple inventory models (e.g., economic order quantity) Simple inventory models (e.g., economic order quantity) Transportation (minimize cost of shipments) Transportation (minimize cost of shipments)
38/47

9. Multiple Goals, Sensitivity Analysis, 9. Multiple Goals, Sensitivity Analysis, What-If Analysis, and Goal Seeking What-If Analysis, and Goal Seeking
Multiple Goals Multiple Goals
Todays management systems are more complex, and one with single goals is Todays management systems are more complex, and one with aasingle goals is rare managers want to attain simultaneousgoals some of which may conflict. goals, rare -- managers want to attain simultaneous goals, some of which may conflict. It is usually necessary to transform multiple-goal problem into singleIt is usually necessary to transform aa multiple-goal problem into aa singlemeasure-of-effectiveness problem before comparing the effects of the solutions measure-of-effectiveness problem before comparing the effects of the solutions (handling multiple goals in LP model goal programming model). (handling multiple goals in aaLP model goal programming model). Methods of handling multiple goals used when working with MSS: Methods of handling multiple goals used when working with MSS: Utility theory Utility theory Goal programming Goal programming Expression of goals as constraints, using LP Expression of goals as constraints, using LP A point system A point system With some methods, the decision maker needs to search the solution space for With some methods, the decision maker needs to search the solution space for an alternative that provides for required attainment of all goals while searching for an alternative that provides for required attainment of all goals while searching for an efficient solution. an efficient solution.
39/47

 9. Multiple Goals, Sensitivity Analysis, What-If Analysis, and Goal Seeking 9. Multiple Goals, Sensitivity Analysis, What-If Analysis, and Goal Seeking

Sensitivity Analysis Sensitivity Analysis

Sensitivity analysis attempts to assess the impact of change on the input data Sensitivity analysis attempts to assess the impact of aachange on the input data or parameters on the proposed solution (result variables). or parameters on the proposed solution (result variables). Sensitivity analysis is important in MSS because it allows flexibility and Sensitivity analysis is important in MSS because it allows flexibility and adaptation to changing conditions and to the requirements of different decisionadaptation to changing conditions and to the requirements of different decisionmaking situation, provides better understanding of the model and the decisionmaking situation, provides aa better understanding of the model and the decisionmaking situation it attempts to describe, and permits the manager to input data in making situation it attempts to describe, and permits the manager to input data in order to increase the confidence in the model. order to increase the confidence in the model. Sensitivity analysis tests relationships such as the following: Sensitivity analysis tests relationships such as the following: The impact of changes in uncontrollable variables and parameters on the The impact of changes in uncontrollable variables and parameters on the outcome variables outcome variables The impact of changes in decision variables and parameters on the outcome The impact of changes in decision variables and parameters on the outcome variables variables The effect of uncertainty in estimating external variables The effect of uncertainty in estimating external variables The effect of different dependent interactions among variables The effect of different dependent interactions among variables The robustness of decisions under changing conditions. The robustness of decisions under changing conditions.
40/47

 9. Multiple Goals, Sensitivity Analysis, What-If Analysis, and Goal Seeking 9. Multiple Goals, Sensitivity Analysis, What-If Analysis, and Goal Seeking
Sensitivity Analysis Sensitivity Analysis

Sensitivity analysis are used for: Sensitivity analysis are used for: Revising models to eliminate too-large sensitivities Revising models to eliminate too-large sensitivities Adding details about sensitive variables or scenarios Adding details about sensitive variables or scenarios Obtaining better estimates of sensitive external variables Obtaining better estimates of sensitive external variables Accepting and using the sensitive (and hence vulnerable) real world, leading Accepting and using the sensitive (and hence vulnerable) real world, leading to continuous and close monitoring of actual results to continuous and close monitoring of actual results The robustness of decisions under changing conditions. The robustness of decisions under changing conditions.

Automatic Sensitivity Analysis Automatic Sensitivity Analysis

ASA is performed in standard quantitative model implementation such as LP. ASA is performed in standard quantitative model implementation such as LP. Automatic sensitivity analysis is usually limited to one change at time, and Automatic sensitivity analysis is usually limited to one change at aa time, and only for certain variables. It is very powerful because of its ability to establish only for certain variables. It is very powerful because of its ability to establish ranges and limits very fast. ranges and limits very fast.
41/47

 9. Multiple Goals, Sensitivity Analysis, What-If Analysis, and Goal Seeking 9. Multiple Goals, Sensitivity Analysis, What-If Analysis, and Goal Seeking

What-If Analysis What-If Analysis What-if analysis is structured as What will happen to the solution if an input What-if analysis is structured as What will happen to the solution if an input
variable, an assumption, or parameter value is changed? variable, an assumption, or aaparameter value is changed?
a = 2.50 x y f(x,y) g(x,y) h(x,y) $100.00 $200.00 $600.00 $325.00 $215.00 $110.00 $220.00 $660.00 $357.50 $236.50 b = 1.75 $121.00 $242.00 $726.00 $393.25 $260.15 $133.10 $266.20 $798.60 $432.58 $286.17

opportunity to change their answers to some of the systems questions, and opportunity to change their answers to some of the systems questions, and aa revised recommendation is found. revised recommendation is found.
42/47

What-if analysis is common in expert systems Users are given the systems. What-if analysis is common in expert systems. Users are given the

 9. Multiple Goals, Sensitivity Analysis, What-If Analysis, and Goal Seeking 9. Multiple Goals, Sensitivity Analysis, What-If Analysis, and Goal Seeking

Goal Seeking Goal Seeking

level of an output (goal). It represents backward solution approach. level of an output (goal). It represents aabackward solution approach.

Goal seeking calculates the values of the inputs necessary to achieve desired Goal seeking calculates the values of the inputs necessary to achieve aadesired
The following are some The following are some examples of goalseeking seeking: examples of goal seeking: What annual budget is needed What annual budget is needed for an annual growth rate of for an annual growth rate of 15% by 2009? 15% by 2009? How many nurses are needed How many nurses are needed to reduce the average waiting to reduce the average waiting time of patient in the time of aa patient in the emergency room to less than 10 emergency room to less than 10 minutes? minutes?
43/47

10. Problem-Solving Search Methods 10. Problem-Solving Search Methods

Search Methods are used in the choice phase of problem solving, and include Search Methods are used in the choice phase of problem solving, and include analyticaltechniques algorithms blindsearching and heuristicsearching techniques, algorithms, searching, searching. analytical techniques, algorithms, blind searching, and heuristic searching. Formal Search Approaches: Formal Search Approaches:
Optimization (Analytical) Search Process
Generate improved solutions or get the best solution directly All possible solutions are checked Check only some alternatives Only promising solutions are considered

Stop Testing
Stop when no improvement is possible Comparisons: Stop when all alternatives are checked Comparisons, simulation: Stop when solution is good enough Stop when solution is good enough

Solution
Optimal (best) Optimal (best) Best among alternatives checked Good enough
44/47

Search Approaches

Blind Search

Complete enumeration (exhaustive) Partial search

Heuristics

 10. Problem-Solving Search Methods 10. Problem-Solving Search Methods

Analytical Techniques Analytical Techniques

Analytical Techniques use mathematical formulas to derive an optimal solution Analytical Techniques use mathematical formulas to derive an optimal solution directly to predict certain result. Analytical techniques are used mainly for directly to predict aa certain result. Analytical techniques are used mainly for solving structured problems, usually of tactical or operational nature, in areas solving structured problems, usually of aa tactical or operational nature, in areas such as resources allocation or inventory management. Blind or heuristic search such as resources allocation or inventory management. Blind or heuristic search approaches are generally employed to solve more complex problems. approaches are generally employed to solve more complex problems.
Start

Algorithms Algorithms
Analytical Techniques Analytical Techniques may use algorithms to may use algorithms to increase the efficiency of increase the efficiency of the search. An algorithm the search. An algorithm is step-by-step search is step-by-step search process for obtaining an process for obtaining an optimal solution. optimal solution.

Is improvement possible in proposed solution? Improve solution. generate a new proposed solution.

No

Solution is optimal

End

Yes
45/47

 10. Problem-Solving Search Methods 10. Problem-Solving Search Methods

Blind Searching Blind Searching

In conducting search, description of desired solution may be given. This In conducting aasearch, aadescription of aadesired solution may be given. This is called goal A set of possible steps leading from initial conditions to the goal goal. is called aagoal. A set of possible steps leading from initial conditions to the goal is called the search steps Problem solving is done bye searching through the steps. is called the search steps. Problem solving is done bye searching through the possible solutions. The first of these search methods is blind searching The searching. possible solutions. The first of these search methods is blind searching. The second is heuristicsearching searching. second is heuristic searching. Blind search techniques are arbitrary search approaches that are not guided. Blind search techniques are arbitrary search approaches that are not guided. There are two types of blind searches: There are two types of blind searches: complete enumeration an optimal solution is discovered, aacomplete enumeration an optimal solution is discovered, an incomplete, or partial, search until good-enough solution is found. an incomplete, or partial, search until aagood-enough solution is found.

Heuristic Searching Heuristic Searching

Heuristics are the informal, judgment knowledge of an application area that Heuristics are the informal, judgment knowledge of an application area that constitute the rules of good judgment in the field. Through domain knowledge, constitute the rules of good judgment in the field. Through domain knowledge, they guide the problem-solving process. Heuristic programming is the process of they guide the problem-solving process. Heuristic programming is the process of using heuristics in problem solving. using heuristics in problem solving.
46/47

Efraim Turban, Jay E. Aronson, Ting-Peng Liang, Ramesh Sharda Efraim Turban, Jay E. Aronson, Ting-Peng Liang, Ramesh Sharda Pearson Pretince Hall, New Jersey, 2007 Pearson Pretince Hall, New Jersey, 2007

Decision Support and Business Intelligence Systems Decision Support and Business Intelligence Systems

End of 10.

References References
1. Baker, D., Bridges, D., Hunter, R., Johnson, G., Krupa, J., Murphy, J. and Sorenson, K. (2002) Guidebook to 1. Baker, D., Bridges, D., Hunter, R., Johnson, G., Krupa, J., Murphy, J. and Sorenson, K. (2002) Guidebook to Decision- Making Methods, WSRC-IM-2002-00002, Department of Energy, USA. http://emiDecision- Making Methods, WSRC-IM-2002-00002, Department of Energy, USA. http://emiweb.inel.gov/Nissmg/Guidebook_2002.pdf web.inel.gov/Nissmg/Guidebook_2002.pdf 2. Averill M. Law, W. David Kelton, Simulation Modeling and Analysis, 3/ed ,0070592926, 2000 2. Averill M. Law, W. David Kelton, Simulation Modeling and Analysis, 3/ed ,0070592926, 2000 3. Alvin E. Roth, Marilda A. Oliveira Sotomayor, Two-sided matching: study in game-theoretic modeling and 3. Alvin E. Roth, Marilda A. Oliveira Sotomayor, Two-sided matching: aastudy in game-theoretic modeling and analysis, analysis, 4. Archishman Chakraborty, Alessandro Citanna, Michael Ostrovsky, Two-sided matching with interdependent 4. Archishman Chakraborty, Alessandro Citanna, Michael Ostrovsky, Two-sided matching with interdependent values, Schulich School of Business, York University, Toronto ON M3J 1P3, Canada, values, Schulich School of Business, York University, Toronto ON M3J 1P3, Canada, 5. Modeling and Analysis of Dynamic Systems, Modeling and Analysis of Dynamic Systems, 3rd Edition 5. Modeling and Analysis of Dynamic Systems, Modeling and Analysis of Dynamic Systems, 3rd Edition 6. Charles M. Close (Rensselaer Polytechnic Institute), Dean K. Frederick (Unified Technologies, Inc.), Jonathan 6. Charles M. Close (Rensselaer Polytechnic Institute), Dean K. Frederick (Unified Technologies, Inc.), Jonathan C. Newell (Rensselaer Polytechnic Institute), ISBN: 978-0-471-39442-6, August 2001, 2002 C. Newell (Rensselaer Polytechnic Institute), ISBN: 978-0-471-39442-6, August 2001, 2002 7. Sudipto Banerjee Alan Gelfand and Bradley Carlin, Hierarchical Modeling and Analysis for Spatial 7. Sudipto Banerjee , ,Alan EE. .Gelfand , ,and Bradley PP. .Carlin, Hierarchical Modeling and Analysis for Spatial Data, Chapman and Hall/CRC 2004, Print ISBN: 978-1-58488-410-1, eBook ISBN: 978-0-203-48780-8 Data, Chapman and Hall/CRC 2004, Print ISBN: 978-1-58488-410-1, eBook ISBN: 978-0-203-48780-8

. . . C10 / 6.12.2012

47/47