Ch. 14 Decision Analysis Flashcards by Anthony LoBasso

Primary problems in the decision making process include:

Uncertainty regarding the future
Lack of data
Conflicting objectives

How well did you know this?

Not at all

Perfectly

A choiceamong several alternatives

Decision

How well did you know this?

Not at all

Perfectly

A course of action intended to solve a problem

Alternative

How well did you know this?

Not at all

Perfectly

Represents various factors that are important to the decision maker and influenced by alternatives

Criteria

How well did you know this?

Not at all

Perfectly

Correspond to future events that are not under the decision makers control

State of Nature

How well did you know this?

Not at all

Perfectly

The Three types of decision making environments:

Decision-making under 1) risk, 2) certainty, 3) uncertainty

How well did you know this?

Not at all

Perfectly

The list of possible future events is known, and the probability of each event occurring can be assigned a probability

Decision-making under risk

How well did you know this?

Not at all

Perfectly

The list of possible future events is known with complete certitude

Decision-making under certainty

How well did you know this?

Not at all

Perfectly

The list of possible future events is known, but the associated probabilities are unknown

Decision-making under uncertainty

How well did you know this?

Not at all

Perfectly

A tabular rendering of the various alternatives and states that reports the corresponding outcomes.
Summarizes the final outcome (or payoff) for each decision alternative under each possible state of nature.

Payoff Table (Matrix)

How well did you know this?

Not at all

Perfectly

The two basic structural components of a payoff table are?

States of nature and Alternatives

How well did you know this?

Not at all

Perfectly

In a payoff table -

columns equal?
rows equal?
Bottom row equals?

Columns = the decision options (alternatives)
Rows = states of nature
Bottom row = expected value

How well did you know this?

Not at all

Perfectly

What is a probabilities range in value between?

zero and one

How well did you know this?

Not at all

Perfectly

represent the likelihood of occurrence based on historical data
examples: weather and gaming

Objective probabilities

How well did you know this?

Not at all

Perfectly

A process for selecting from several options based on facts, observations, and data

Objective decision-making

How well did you know this?

Not at all

Perfectly

when there is no historical data available. It is PERSONAL

ex. Interviews with experts in the field

Subjective probabilities

How well did you know this?

Not at all

Perfectly

A process for selecting from several options based on assumptions, beliefs, and opinions

Subjective decision-making

How well did you know this?

Not at all

Perfectly

A function that maps an observation to an appropriate action

Decision rule

How well did you know this?

Not at all

Perfectly

The weighted average of a variable where the weights are the respective probabilities associated with the various outcomes

Expected value

How well did you know this?

Not at all

Perfectly

The decision rule can be divided into two methods:

Probabilistic Methods

Non Probabilistic Methods

How well did you know this?

Not at all

Perfectly

Those that assume the probabilities of occurrence can be assigned to the states of nature in a decision problem

Probabilistic Methods

How well did you know this?

Not at all

Perfectly

Those that DO NOT assume the probabilities of occurrence can be assigned to the states of nature in a decision problem

Non Probabilistic Methods

How well did you know this?

Not at all

Perfectly

You use the following when probabilities are what?
Expected Monetary Value Decision Rule (EMV)
Expected Opportunity Loss (EOL)
Expected Value of Perfect Information (EVPI)

PROBABILITIES ARE KNOWN

How well did you know this?

Not at all

Perfectly

You use the following when probabilities are what?
Maximax decision rule
Maximin decision rule
Minimax Regrest decision rule

PROBABILITIES ARE UNKNOWN

How well did you know this?

Not at all

Perfectly

Select the alternative with the largest expected monetary value (EMV)

Expected Monetary Value Decision Rule

for a given decision alternative, this indicates the average monetary outcome of a decision if it was repeated a large number of times

Expected Monetary Value (EMV)

The maximum one is willing to pay for perfect information. | Also, The decision with the smallest EOL will always have the largest EMV

Expected Opportunity Loss (EOL)

expected value obtained with perfect information minus maximum EMV (without sample information)

Expected Value of Perfect Information (EVPI)

Identify the maximum payoff for each alternative and choose the alternative with the largest maximum payoff. Also says nature will always be on our side regardless of the decision

Maximax decision rule

Identify the minimum payoff for each alternative and choose the alternative with the largest minimum payoff. Also says nature will always be against us regardless of the decision

Maximin decision rule

Compute the possible loss for each alternative under each state of nature; identify the maximum possible regret loss for each alternative; choose the alternative with the smallest maximum loss

Minimax Regret decision rule

A modelling system that supports the decision-making process ``` Steps include: 1 Problem Statement 2 Model Selection 3 Evaluate alternatives 4 Select decision 5 Implement ```

Decision support system

A graphical decision support system used to analyze problems involving risk. Arranges decision alternatives and states of nature as prescribed by the problem situation. Usually more effective multi decision type applications

Decision trees

How do you evaluate a decision tree?

RIGHT TO LEFT

when dealing with decision trees, a process to determine the largest EMV

Rolling Back

when dealing with decision trees, you do this by adding double vertical lines to the suboptimal event branch, and it represents an optimal alternative at a decision node.

Pruning

Decision problems often involve two or more conflicting criterion and two or more alternatives: Ex Job Offers involve - salary, location and career potential Car buying - price, warranty, mileage, safety Sometimes these criteria conflict with one another.

Multicriteria Decision-Making

The best when there are multiple decision criteria

Multi Attribute Analysis

Typically each alternative and criterion pair are ranked on a scale. Weights are assigned to each criterion. A weighted average is computed for each alternative. The alternative with the largest weighted average is selected

Multicriteria scoring model

provide an effective way of graphically summarizing numerous alternatives in a multicriteria scoring model

Radar charts

Provides a more structured approach for determining scores and weights for the multicriteria scoring model Especially helpful in focusing attention and discussion on the important aspects of a problem in group decision making environments.

Analytical Hierarchy Process (AHP)

Step 1 Create a pairwise comparison matrix for each alternative on each criterion Step 2 Normalizing the comparisons Step 3 Apply consistency in the preference ratings given in the pairwise comparison matrix Step 4 Obtain scores for the remaining criteria Step 5 Obtain criterion weights Step 6 Implement the scoring model

Steps for the Analytical Hierarchy Process (AHP)

Problems where a multiple decisions must be made to get to the final outcome.

MULTI STAGE DECISION PROBLEMS

A graph or a tree that shows the chances associated with possible outcomes. It is an effective tool for breaking an EMV into its component parts and communicating information about the actual outcomes that can occur as the result of various decisions.

Risk Profile

A process for making decisions where there is more than one criterion

Multifactor modelling

Includes often available information about the possible outcomes of decisions before the decisions are made. This sample information allows us to refind probability estimates associated with various outcomes. The basic tradeoff is the cost of the sample information versus its worth. Pay less for imperfect sample information.

Sample Information in Decision-Making

Examining the impact of a change in one or more of the model coefficients on the decision or decisions. If a small change in one of the model coefficients results in a different decision then the data needs to be investigated in more detail. Is important because most model coefficients are estimates Can be used to assess the impact of changes in both the probability estimates, payoff values, and factor rankings

Sensitivity Analysis

When doing sensitivity analysis, this helps identify the inputs that, if changed, have the greats impact on the EMV. Also helps to identify the areas where sensitivity analysis is most important and prioritize where time and resources should be applied in refining probability and financial estimates represented in the decision tree.

Tornado Charts

a sensitivity analysis technique that allows a decision maker to analyze how the optimal decision strategy changes in response to two simultaneous changes in probability estimates.

Strategy Table

A sensitivity analysis technique to graphically show how the optimal decision strategy changes in response to two simultaneous changes in probability estimates.

Strategy Charts

A matrix that reports indictor probabilities based actual conditions.

Conditional probability table

A process for revising past probabilities using additional prior information.

Bayesian theorem

What are the two inputs required for the Bayesian Theorem?

Model requires two inputs: 1 Prior probabilities 2 Conditional probability table

What are the two outputs required for the Bayesian Theorem?

Model generates two outputs: 1 Marginal probabilities 2 Revised conditional probabilities

Provides a way to incorporate the decision makers attitudes and preferences toward risk and return in the decision analysis process so that the most desirable decision alternative is identified

Utility Theory

The theory assumes that these must be used to translate each of the possible payoffs in a decision problem into a nonmonetary measure known as utility.

Utility Function

Represents the total worth, value or desirability of the outcome of a decision alternative to the decision maker.

Utility of Payoff

Refers to the amount of money that is equivalent in a decision makers mind to a situation that involves uncertainty.

Certainty equivalent

Refers to the EMV that a decision maker is willing to give up (or pay) in order to avoid a risky decision.

Risk Premium

If the decision maker is risk averse, this can be used as an approximation of the decision maker's actual utility function. Must determine a reasonable value for the risk tolerance parameter

Exponential Utility Function

Rigorous definition of problem or issue Comprehensive retrieval of best practices and information (BI) Critical and systematic assessment of evidence using analytics Decision selection and execution Ongoing evaluation of organizational performance

Evidence-Based Decision-Making

``` Factors you need to include when doing this: Customer Preferences Analytics Expertise Customer circumstances ```

Evidence-Based Decision-Making

If two strategies have the same expected profit, select one with the smaller standard deviation

Risk Averting Behavior

What is the probability of exceeding one standard deviation from the mean in a normal distribution?

0.16

``` Which one of the following has the most risk based on the given coefficient of variation? 1 2 3 4 ```

a technique that involves the analysis of a complex decision situation by performing a large number of iterations to determine the probability distribution

Simulation