Slides 8a: Introduction
Decision Analysis Two Topics Decision Theory Framework Decision Trees
Decision Analysis Basic Terms Decision Alternatives States of Nature (eg. Condition of economy or weather) Payoffs ($ outcome of a choice assuming a state of nature) Criteria (i.e. Expected Value)
Decision Analysis Decision Alternatives (“what ifs”) are known States of Nature and their probabilities are known (ex. rainy 20%, partly cloudy 50%, sunny 20%) Outcomes, (referred to as “Payoffs”) are computable under different possible scenarios –each is a combination of a decision alternative and a state of nature
Decision Analysis A set of alternative actions - di You select one of the alternative decisions “What if’s” A set of possible states of nature Only one will be correct, but we don’t know in advance After alternative is selected, only one state of nature occurs which is beyond our control Probabilities may be known are known A set of outcomes and a value for each - rij Each is a combination of an alternative action and a state of nature Payoff - Value can be monetary or otherwise
Decision Analysis Certainty Ignorance Risk Decision Maker knows with certainty what the state of nature will be - only one possible state of nature Ignorance Decision Maker knows all possible states of nature, but does not know probability of occurrence Risk Decision Maker knows all possible states of nature, and can assign probability of occurrence for each state Assumption about natures behavoir
Decision Making Under Certainty
Decision Making Under Ignorance – Payoff Table Kelly Construction Payoff Table (Prob. 8-17) Kelly Construction wants to get in on the boom of student condominium construction. The company must decide whether to purchase enough land to build a 50-, 100-, or 150-unit condominium complex. Many other complexes are currently under construction, so Kelly is unsure how strong demand for its complex will be. If the company is conservative and builds only a few units, it loses potential profits if the demand turns out to be high. On the other hand, many unsold units would also be costly to Kelly. The following table has been prepared, based on three levels of demand.
Decision Making Under Ignorance – Payoff Table Kelly Construction Payoff Table (Prob. 8-17)
Decision Making Under Ignorance LaPlace-Bayes All states of nature are equally likely to occur. Select alternative with best average payoff Maximax Select the strategy with the highest possible return Maximin Select the strategy with the smallest possible loss
LaPlace-Bayes Select Alternative with best average payoff
Maximax: The Optimistic Point of View Select the “best of the best” strategy Evaluates each decision by the maximum possible return associated with that decision (Note: if cost data is used, the minimum return is “best”) The decision that yields the maximum of these maximum returns (maximax) is then selected For “risk takers” Doesn’t consider the “down side” risk Ignores the possible losses from the selected alternative
Maximax Example Kelly Construction
Maximin: The Pessimistic Point of View Select the “best of the worst” strategy Evaluates each decision by the minimum possible return associated with the decision The decision that yields the maximum value of the minimum returns (maximin) is selected For “risk averse” decision makers A “protect” strategy Worst case scenario the focus
Maximin Kelly Construction
Decision Making Under Risk Expected Return (ER) or Expected Value (EV) or Expected Monetary Value (EMV) Select the alternative with the highest (long term) expected return A weighted average of the possible returns for each alternative, with the probabilities used as weights Note that this amount will not be obtained in the short term; the true amount will be for that alternative with the state of nature that actually occurs.
Expected Return
Expected Value of Perfect Information EVPI measures how much better you could do on this decision if you could always know when each state of nature would occur, where: EVUPI = Expected Value Under Perfect Information (also called EVwPI, the EV with perfect information, or EVC, the EV “under certainty”) EVUII = Expected Monetary Value of the best action with imperfect information (also called EMVBest ) EVPI = EVUPI – EVUII EVPI tells you how much you are willing to pay for perfect information (or is the upper limit for what you would pay for additional “imperfect” information!)
Expected Value of Perfect Information To calculate the expected value of perfect information, choose the maximum value for each outcome (column) and multiply it by its respective probability. Then, add the resulting products.
Expected Value of Perfect Information
Using Excel to Calculate EVPI: Formulas View Kelly Construction
The Newsvendor Model A newsvendor can buy the Wall Street Journal newspapers for 40 cents each and sell them for 75 cents. However, he must buy the papers before he knows how many he can actually sell. If he buys more papers than he can sell, he disposes of the excess at no additional cost. If he does not buy enough papers, he loses potential sales now and possibly in the future. Suppose that the loss of future sales is captured by a loss of goodwill cost of 50 cents per unsatisfied customer.
The demand distribution is as follows: P0 = Prob{demand = 0} = 0.1 P1 = Prob{demand = 1} = 0.3 P2 = Prob{demand = 2} = 0.4 P3 = Prob{demand = 3} = 0.2 Each of these four values represent the states of nature. The number of papers ordered is the decision. The returns or payoffs are as follows:
State of Nature (Demand) 0 1 2 3 Decision 0 0 -50 -100 -150 1 -40 35 -15 -65 2 -80 -5 70 20 3 -120 -45 30 105 Payoff = 75(# papers sold) – 40(# papers ordered) – 50(unmet demand) Where 75¢ = selling price 40¢ = cost of buying a paper 50¢ = cost of loss of goodwill
State of Nature (Demand) Now, the ER is calculated for each decision: ER0 = 0(0.1) – 50(0.3) – 100(0.4) – 150(0.2) = -85 ER1 = -40(0.1) + 35(0.3) – 15(0.4) – 65(0.2) = -12.5 ER2 = -80(0.1) – 5(0.3) + 70(0.4) + 20(0.2) = 22.5 ER3 = -120(0.1) – 45(0.3) + 30(0.4) – 105(0.2) = 7.5 State of Nature (Demand) 0 1 2 3 Decision 0 0 -50 -100 -150 -85 1 -40 35 -15 -65 -12.5 2 -80 -5 70 20 22.5 3 -120 -45 30 105 7.5 ER Prob. 0.1 0.3 0.4 0.2 Of these four ER’s, choose the maximum, and order 2 papers
ER(current) = 22.5 EVPI = 59.5 – 22.5 = 37.0 cents State of Nature 0 1 2 3 Decision 0 0 -50 -100 -150 1 -40 35 -15 -65 2 -80 -5 70 20 3 -120 -45 30 105 Prob. 0.1 0.3 0.4 0.2 ER(new) = 0(0.1) + 35(0.3) + 70(0.4) + 105(0.2) = 59.5 ER(current) = 22.5 EVPI = 59.5 – 22.5 = 37.0 cents
Maximax Criterion: The Maximax criterion is an optimistic decision making criterion. This method evaluates each decision by the maximum possible return associated with that decision. The decision that yields the maximum of these maximum returns (maximax) is then selected.
Maximin Criterion: The Maximin criterion is an extremely conservative, or pessimistic, approach to making decisions. Maximin evaluates each decision by the minimum possible return associated with the decision. Then, the decision that yields the maximum value of the minimum returns (maximin) is selected.
So, using the 3 criteria, we made the following decisions regarding the newsvendor data: Criteria Decision Maximin Cash Flow Order 1 paper Expected Return Order 2 papers Maximax Cash Flow Order 3 papers