Presentation on theme: "Introduction to Decision Analysis"— Presentation transcript:
1Introduction to Decision Analysis CD-ROM Chapter 17Introduction to Decision Analysis
2Chapter 17 - Chapter Outcomes After studying the material in this chapter, you should be able to:Describe the decision-making environments of certainty and uncertainty.Construct both a payoff table and an opportunity loss table.Define the expected value criterion.Apply the expected value criterion in decision situations.Compute the value of perfect information.
3Chapter 17 - Chapter Outcomes (continued) After studying the material in this chapter, you should be able to:Develop a decision tree and explain how it can aid decision making in an uncertain environment.Discuss the difference between risk seeking and risk avoiding behavior.Construct an individual risk preference function.
4Decision-Making Environments Certainty refers to a decision environment in which the results of selecting each alternative are known before the decision is made.
5Decision-Making Environments Uncertainty refers to a decision environment in which the decision maker does not know what outcome will occur when an alternative is selected.
6Decision-Making Environments The goal of decision analysis is to focus on making good decisions, which in the long run should result in an increased number of good outcomes.
7Decision CriteriaThe states of nature are the possible outcomes in a decision situation over which the decision maker has no control.
8Decision CriteriaA payoff is the outcome (profit or loss) for any combination of alternative states of nature. The outcomes of all possible combinations of alternatives and states of nature constitute a payoff table.
10Decision CriteriaThe maximax criterion is an optimistic decision criterion for dealing with uncertainty without using probability. For each option, the decision maker finds the maximum possible payoff and then selects the option with the greatest maximum payoff.
11Decision CriteriaThe maximin criterion is a pessimistic (conservative) decision criterion for dealing with uncertainty without using probability. For each option, the decision maker finds the minimum possible payoff and then selects the option with the greatest minimum payoff.
12Decision CriteriaThe opportunity loss is the difference between the actual payoff that occurs for a decision and the optimal payoff for the same decision.
13Decision CriteriaThe minimax regret criterion is a decision criterion that considers the costs of selecting the “wrong” alternative. For each sate of nature, the decision maker finds the difference between the best payoff and each other alternative and uses these values to construct an opportunity-loss table. The decision maker then selects the alternative with the minimum opportunity loss (or regret).
15Decision Criteria (Table 17-4) Fisher Fabrication Maximum Regret Table
16Decision CriteriaThe expected-value criterion is a decision criterion that employs probability to select the alternative that will produce the greatest average payoff or minimum average loss.
17Decision Criteria EXPECTED VALUE where: xi = The ith outcome of the specified alternative measured in some units, such as dollarsP(xi) = The probability of outcome xi occurringk = number of potential outcomesand:
18CLASSICAL PROBABILITY ASSESSMENT Decision CriteriaCLASSICAL PROBABILITY ASSESSMENT
19RELATIVE FREQUENCY OF OCCURRENCE PROBABILITY where: Decision CriteriaRELATIVE FREQUENCY OF OCCURRENCE PROBABILITYwhere:
22Decision-Tree Analysis A decision tree is a diagram that illustrates the correct ordering of actions and events in a decision-analysis problem. Each act or event is represented by a node on the decision tree.
31Risk Preference Attitudes A risk-averse attitude refers to the preference for risk such that the decision maker could select an alternative with a lower expected payoff in order to avoid the possibility of an undesirable outcome.
32Risk Preference Attitudes Certainty equivalent is the value that would make a decision maker indifferent between taking an uncertain gamble versus receiving that value instead of taking the gamble.
33Risk Preference Attitudes A risk-seeking attitude refers to the preference for risk such that the decision maker could select an alternative with a lower expected payoff in hopes of achieving an outcome with a more desirable result.
34Risk Preference Attitudes The risk preference function is the graph that describes a decision maker’s preference for risk over the range of possible payoffs.
35Risk Preference Attitudes A standard gamble approach is the approach for assessing risk-preference functions that involves setting up a series of gambles between two payoffs and determining the certainty equivalent for each gamble.
36Risk Preference Attitudes A preference quotient refers to the measure of the relative utility for the outcomes of a decision on a scale between 0.0 and 1.0.
37Risk Preference Attitudes (Figure 17-16) End Valuesq Values$10,0001.00.5Play0.5-$2,0000.0CE = ?Don’t PlayAssessing the Risk-Preference Function: Standard Gamble 1
38Risk Preference Attitudes (Figure 17-17) End Valuesq Values$10,0001.00.5Play0.5$4,0000.5CE = ?Don’t PlayAssessing the Risk-Preference Function: Standard Gamble 2
39Risk Preference Attitudes (Figure 17-18) End Valuesq Values$4,0000.50.5Play0.5-$2,0000.0CE = ?Don’t PlayAssessing the Risk-Preference Function: Standard Gamble 3
40Risk Preference Attitudes Risk premium is the difference between the expected value of an event and the certainty equivalent. The risk premium will be zero for a risk-neutral decision maker, positive for a risk-averse decision maker, and negative for a risk-seeking decision maker.
41Risk Preference Attitudes (Figure 17-19) 0.750.500.25-$2,000$0$2,000$4,000$6,000$8,000$10,000Risk-Neutral Preference Function
42Risk Preference Attitudes (Figure 17-23) 0.750.500.25-$2,000$0$2,000$4,000$6,000$8,000$10,000Risk-Averse Preference Function
43Risk Preference Attitudes (Figure 17-26) 0.750.500.25-$2,000$0$2,000$4,000$6,000$8,000$10,000Risk-Seeking Preference Function