Decision Analysis (1)

  • Upload
    myra

  • View
    213

  • Download
    0

Embed Size (px)

Citation preview

  • 7/28/2019 Decision Analysis (1)

    1/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-1

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Decision Analysis(Theory)

  • 7/28/2019 Decision Analysis (1)

    2/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-2

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Introduction

    Decision theory is an analytical

    and systematic way to tackle

    problems

    A good decision is based on

    logic.

  • 7/28/2019 Decision Analysis (1)

    3/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-3

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    The Six Steps inDecision Theory

    1) Clearly define the problem at

    hand

    2) List the possible alternatives3) Identify the possible outcomes

    4) List the payoff or profit of

    each combination ofalternatives and outcomes

    5) Select one of the mathematical

    decision theory models6) Apply the model and make

    your decision

  • 7/28/2019 Decision Analysis (1)

    4/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-4

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Decision Tablefor Thompson Lumber

    State of Nature

    Alternative FavorableMarket

    UnfavorableMarket

    Construct alarge plant

    $200,000 -$180,000

    Construct asmall plant

    $100,000 -$20,000

    Do nothing $0 0

  • 7/28/2019 Decision Analysis (1)

    5/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-5

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Types of Decision-Making Environments

    Type 1: Decision-making under certainty

    decision-makerknows with certainty the

    consequences of every alternative or decision choice

    Type 2: Decision-making under risk

    The decision-makerdoes know the probabilities of

    the various outcomes

    Type 3: Decision-making under uncertainty

    The decision-makerdoes not know the probabilities

    of the various outcomes

  • 7/28/2019 Decision Analysis (1)

    6/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-6

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Decision-MakingCertainty

    Knows with certainty the result

    of every alternative

  • 7/28/2019 Decision Analysis (1)

    7/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-7

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Decision-MakingUnder Risk

    Expected Monetary Value

    nature.ofstagesofnumbernwhere

    )(*SPayoffative)EMV(Altern1

    j

    j

    n

    j

    SP

  • 7/28/2019 Decision Analysis (1)

    8/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-8

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Decision Table(Matrix)

    for Thompson Lumber

    Favorable

    Market

    Unfavorable

    MarketAlternative State of Nature

    Construct a

    large plant

    $200,000 -$180,000 $10,000

    Construct asmall plant

    $100,000 -$20,000 $40,000

    Do nothing $0 0

    0.50 0.50

  • 7/28/2019 Decision Analysis (1)

    9/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-9

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Expected Value ofPerfect Information

    (EVPI)

    EVPIplaces an upper bound on

    what one would pay for

    additional information

    EVPIis the expected value with

    perfect information minus themaximum EMV

  • 7/28/2019 Decision Analysis (1)

    10/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-10

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Expected Value WithPerfect Information

    (EV | PI)

    nature.ofstatesofnumbern

    )P(S*nature)ofstateforoutcome(BestPI|EVn

    1j

    j

  • 7/28/2019 Decision Analysis (1)

    11/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-11

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Expected Value ofPerfect Information

    EVPI=EV|PI- maximumEMV

  • 7/28/2019 Decision Analysis (1)

    12/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-12

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Expected Value ofPerfect Information

    State of Nature

    Alternative Favorable

    Market

    Unfavorable

    Market

    EMV

    Construct alarge plant

    $200,000

    Construct a

    small plant

    $40,000

    Do Nothing $0

    0.50 0.50

  • 7/28/2019 Decision Analysis (1)

    13/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-13

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Expected Value ofPerfect Information

    EVPI= expected value with perfect

    information - max(EMV)

    = $200,000*0.50 + $0*0.50 - $40,000

    = $60,000

  • 7/28/2019 Decision Analysis (1)

    14/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-14

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Expected OpportunityLoss

    EOL is the cost of not picking

    the best solution

    EOL = Expected Regret

  • 7/28/2019 Decision Analysis (1)

    15/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-15

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Computing EOL - TheOpportunity Loss Table

    State of Nature

    ternative Favorable Market

    ($)

    Unfavorabl

    Market ($)

    Large Plant 200,000 - 200,000 0 - (-180,000

    Small Plant 200,000 - 100,000 0 -(-20,000

    Do Nothing 200,000 - 0 0-0

    Probability 0.50 0.50

  • 7/28/2019 Decision Analysis (1)

    16/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-16

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    The Opportunity LossTable continued

    State of Nature

    lternative FavorableMarket

    UnfavorablMarket

    Large Plant 0 $180,000

    Small Plant $100,000 $20,000

    Do Nothing $200,000 0

    Probability 0.50 0.50

  • 7/28/2019 Decision Analysis (1)

    17/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-17

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    The Opportunity LossTable - continued

    lternative EOL

    Large Plant (0.50)*$0 +(0.50)*($180,000)

    $90,000

    Small Plant (0.50)*($100,000)+ (0.50)(*$20,000)

    $60,000

    Do Nothing (0.50)*($200,000)

    + (0.50)*($0)

    $100,000

  • 7/28/2019 Decision Analysis (1)

    18/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-18

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Sensitivity Analysis

    EMV(Large Plant) = $200,000P- (1-

    P)$180,000

    EMV(Small Plant) = $100,000P-

    $20,000(1-P)

    EMV(Do Nothing) = $0P+ 0(1-P)

  • 7/28/2019 Decision Analysis (1)

    19/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-19

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Sensitivity Analysis -continued

    -200000

    -150000

    -100000

    -500000

    50000

    100000

    150000

    200000

    250000

    0 0.2 0.4 0.6 0.8

    Values of P

    EMV

    Values

    Point 1 Point 2Small Plant

    Large Plant EMV

  • 7/28/2019 Decision Analysis (1)

    20/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-20

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Decision MakingUnder Uncertainty

    Maximax

    Maximin

    Equally likely (Laplace)

    Criterion of Realism

    Minimax

  • 7/28/2019 Decision Analysis (1)

    21/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-21

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Decision MakingUnder Uncertainty

    Maximax - Choose the alternativewith the maximum output

    State of Nature

    Alternative FavorableMarket

    UnfavorableMarket

    Construct alarge plant

    200,000 -180,000

    Construct asmall plant

    100,000 -20,000

    Do nothing 0 0

    Probability 0.50 0.50

  • 7/28/2019 Decision Analysis (1)

    22/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-22

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Decision MakingUnder Uncertainty

    Maximin - Choose the alternative withthe maximum minimum output

    State of Nature

    Alternative FavorableMarket

    UnfavorableMarket

    Construct a

    large plant

    200,000 -180,000

    Construct a

    small plant

    100,000 -20,000

    Do nothing 0 0

    Probabilities

  • 7/28/2019 Decision Analysis (1)

    23/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-23

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Decision MakingUnder Uncertainty

    Equally likely (Laplace) - Assume all

    states of nature to be equally

    likely, choose maximum Average

    States of Nature

    Alternative Favorable

    Market

    Unfavorable

    Market

    Avg.

    Construct

    Large Plant

    $200,000 -$180,000 10,000

    Construct

    mall plant

    100,000 -20,000 40,000

    Do nothing 0 0 0

  • 7/28/2019 Decision Analysis (1)

    24/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-24

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Decision MakingUnder Uncertainty

    Criterion of Realism (Hurwicz):

    CR = *(row max) + (1-)*(row min)

    State of Nature

    Alternative Favorable

    Market

    Unfavorable

    Market

    CR

    Constructarge plant $200,000 -180,000 124,000

    Construct

    small plant

    $100,000 -20,000 76,000

    Do nothing 0 0 0

    0.80 0.20

  • 7/28/2019 Decision Analysis (1)

    25/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-25

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Decision MakingUnder Uncertainty

    Minimax - choose the alternative withthe minimum maximum OpportunityLoss

    States of Nature

    Alternative FavorableMarket

    Unfavorable

    MarketMax

    Construct a

    arge plant

    0$ $180,000 $180,000

    Construct a

    mall plant

    $100,000 20,000 100,000

    Do nothing 200,000 0 200,000

    Regret Matrix

  • 7/28/2019 Decision Analysis (1)

    26/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-26

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Marginal Analysis -Discrete Distributions

    Steps using Normal Distributions:

    Determine the value forP.

    Locate P on the normal distribution. For agiven area under the curve, we findZfrom

    thestandard Normal table.

    Using we can now

    solve forX*

    sm

    *

    XZ

    MPML

    ML

    P +

  • 7/28/2019 Decision Analysis (1)

    27/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-27

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Joes NewsstandExample A

    ML = 4

    MP= 6

    m= Average demand = 50

    papers per day

    s= Standard deviation of

    demand = 10

  • 7/28/2019 Decision Analysis (1)

    28/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-28

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Joes Newsstand

    Example A - continued

    Step 1:

    Step 2: Look in the Normal tableforP= 0.6 (i.e., 1 0.4)

    .

    40.064

    4

    +

    +

    MPML

    MLp

    newspapers53or52.5500.25*10X

    or

    10

    50X0.25Z

    *

    *

    +

    -

  • 7/28/2019 Decision Analysis (1)

    29/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-29

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Joes Newsstand

    Example A continued

  • 7/28/2019 Decision Analysis (1)

    30/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-30

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Joes NewsstandExample B

    ML = 8

    MP= 2

    m= Average demand = 100

    papers per day

    s= Standard deviation ofdemand = 10

  • 7/28/2019 Decision Analysis (1)

    31/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-31

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Joes Newsstand

    Example B - continued

    Step 1:

    Step 2:Z= -0.84 for an areaof 0.80

    and

    or:

    80.028

    8

    +

    +

    MPML

    MLp

    10

    1000X0.84

    *-

    -

    newspapers92or91.6

    1000.84(10)X*

    +- 0

  • 7/28/2019 Decision Analysis (1)

    32/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-32

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

    Joes Newsstand

    Example B continued

  • 7/28/2019 Decision Analysis (1)

    33/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-33

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

  • 7/28/2019 Decision Analysis (1)

    34/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-34

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

  • 7/28/2019 Decision Analysis (1)

    35/36

    To accompany Quantitative Analysis

    for Management, 8e

    by Render/Stair/Hanna3-35

    2003 by Prentice Hall, Inc.

    Upper Saddle River, NJ 07458

  • 7/28/2019 Decision Analysis (1)

    36/36