Scribd
Upload
16 views5 pages

Decision Analysis in Operations Research

The document covers decision analysis in operations research, detailing decision environments of certainty, uncertainty, and risk, along with various decision criteria. It explains how to construct payoff and opportunity-loss tables, and outlines decision-making steps and criteria, including non-probabilistic and probabilistic approaches. Examples illustrate the application of different decision-making criteria such as maximax, maximin, realism, equally likely, and minimax regret.

Uploaded by

eliamugini444
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
16 views5 pages

Decision Analysis in Operations Research

The document covers decision analysis in operations research, detailing decision environments of certainty, uncertainty, and risk, along with various decision criteria. It explains how to construct payoff and opportunity-loss tables, and outlines decision-making steps and criteria, including non-probabilistic and probabilistic approaches. Examples illustrate the application of different decision-making criteria such as maximax, maximin, realism, equally likely, and minimax regret.

Uploaded by

eliamugini444
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

INSTITUTE OF ACCOUNTANCY ARUSHA

MODULE: OPERATIONS RESEARCH - AFU07437 / BMU07424 / SSU07404

TOPIC: DECISION ANALYSIS

Lecture 1

Objectives:
After completing this topic, you should be able to:
􀂄Describe the decision environments of certainty, uncertainty and risk.
􀂄Construct a payoff table and an opportunity-loss table.
􀂄Use non-probabilistic and probabilistic criteria for decision making.
􀂄Define and apply the expected value criterion for decision making.
􀂄Compute the value of perfect information.
􀂄Develop and use decision trees for decision making.

Decision Making Overview

Decision Making

Decision Environment Decision Criteria

Certainty Non-probabilistic

Uncertainty
Probabilistic

Risk

The Decision Making Environment

Certainty: The results of decision alternatives are known. This means that the decision
maker knows with certainty the consequences of every alternative or decision choice.
Naturally, the choice would be the alternative that will result in the best outcome.

Example:
Must print 10,000 color brochures.
Offset press A: TZS 200,000 fixed cost + TZS 250 per page
Offset press B: TZS 500,000 fixed cost + TZS 200 per page

Uncertainty: The outcome that will occur after a choice is unknown. This means that
there are several outcomes for each alternative, and the decision maker does not know
the probabilities of the various outcomes. Non-probabilistic decision criteria are applied
under the environment of uncertainty.

Example:

1
You must decide to buy an item now or wait. If you buy now the price is TZS 200,000. If
you wait the price may drop to TZS 150,000 or rise to TZS 220,000. There also may be
a new model available later with better features.

Risk: There are several possible outcomes for each alternative, and the decision maker
knows the probability of occurrence of each outcome. Probabilistic decision criteria are
applied under the environment of risk.

The Decision Making Criteria

Non-probabilistic Decision Criteria:Five decision rules that can be applied if the


probabilities of uncertain events are not known.

􀂄maximax criterion
􀂄maximin criterion
􀂄realism (Hurwicz) criterion
􀂄equally likely (laplace) criterion
􀂄minimax regret criterion

Probabilistic Decision Criteria: Consider the probabilities ofuncertain events and


select analternative to maximize theexpected payoff or minimize theexpected loss. Two
decision rules that can be applied if the probabilities of uncertain events are known.

􀂄maximize expected value


􀂄minimize expected opportunity loss

Steps in Decision Making

􀂄Clearly define the problem.


􀂄List all possible alternatives.
􀂄Identify all possible outcomes for each alternative.
􀂄Identify the payoff for each alternative and outcome combination.
􀂄Use a decision modeling technique or criteria to choose an alternative.

A Payoff Table:

A payoff table shows alternatives, states of nature, and payoffs.

Investment Choice Profit in TZS 1,000,000’s


(Alternatives) (States of Nature)
Strong Economy Stable Economy Weak Economy
Large Factory 200 50 -120
Medium Factory 90 120 -30
Small Factory 40 30 20

Non-probabilistic Decision Criteria:

Maximax Solution:
2
The maximax criterion (an optimistic approach):
1. For each option, find the maximum payoff.
2. Choose the option with the greatest maximum payoff.

In the payoff table above, the maximum profit for each alternative is given below:
Large Factory = TZS 200,000,000
Medium Factory = TZS 120,000,000
Small Factory = TZS 40,000,000

The greatest maximum profit is therefore TZS 200,000,000.


The decision under the Maximax criterion is to choose Large Factory.

Maximin Solution:

The maximin criterion (a pessimistic approach):


1. For each option, find the minimum payoff
2. Choose the option with the greatest minimum payoff

In the payoff table above, the minimum profit for each alternative is given below:
Large Factory = (TZS 120,000,000)
Medium Factory = (TZS 30,000,000)
Small Factory = TZS 20,000,000

The greatest minimum profit is therefore TZS 20,000,000.


The decision under the Maximin criterion is to choose Small Factory.

Realism (Hurwicz) Solution:

The realism criterion:


1. Determine the coefficient of realism or index of optimism (). Note 0  1,
 = 0 Maximin and  = 1Maximax.
2. For each option, find the weighted average (= x maximum payoff + (1-) x
minimum payoff)
2. Choose the option with the greatest weighted average payoff.

In the payoff table above, the weighted average for each alternative is given below ( =
0.6):
Large Factory = 0.6 x 200,000,000 + 0.4 x (-120,000,000) = TZS 72,000,000
Medium Factory = 0.6 x 120,000,000 + 0.4 x (-30,000,000) = TZS 60,000,000
Small Factory = 0.6 x 40,000,000 + 0.4 x 20,000,000 = TZS 32,000,000

The greatest payoff is therefore TZS 72,000,000.


The decision under the Realism criterion is to choose Large Factory.

Equally Likely Solution:

The equally likely criterion (also known as laplace criterion) makes the assumption that
any of the states of nature could occur, but does not give preference to any one.

3
Calculate the weighted average for each alternative and then choose the best
payoff.

In the payoff table above, the weighted average for each alternative is given below:
Large Factory = 0.33 x 200,000,000 + 0.33 x 50,000,000 +0.33 x (-120,000,000) =
TZS 42,900,000
Medium Factory = 0.33 x 90,000,000 + 0.33 x 120,000,000 + 0.33 x (-30,000,000) =
TZS 60,000,000
Small Factory = 0.33 x 40,000,000 + 0.33 x 30,000,000 + 0.33 x 20,000,000 =
TZS 30,000,000

The greatest weighted average is therefore TZS 60,000,000.


The decision under the Equally Likely criterion is to choose Medium Factory.

Opportunity Loss:

Opportunity loss is the difference between an actual payoff for a decision and the
optimal payoff for that state of nature.

Investment Choice Profit in TZS 1,000,000’s


(Alternatives) (States of Nature)
Strong Economy Stable Economy Weak Economy
Large Factory 200 50 -120
Medium Factory 90 120 -30
Small Factory 40 30 20

The choice “Medium Factory” has payoff 90 for “Strong


Economy”. Given “Strong Economy”, the choice of “Large
Factory” would have given apayoff of 200, or 110 higher.
Therefore the opportunity loss = 110 for this cell.

Now, compute the opportunity loss for all cell to obtain the following “Opportunity Loss”
table:

OpportunityLoss Table:
Investment Choice Opportunity Loss in TZS 1,000,000’s
(Alternatives) (States of Nature)
Strong Economy Stable Economy Weak Economy
Large Factory 0 70 140
Medium Factory 110 0 50
Small Factory 160 90 0

Minimax Regret Solution:

1. For each alternative, find the maximum opportunity loss (or “regret”).
2. Choose the option with the smallest maximum loss.

In the Opportunity Loss table above, the maximum opportunity loss (or “regret”) for each
alternative is given below:
4
Large Factory = TZS 140,000,000
Medium Factory = TZS 110,000,000
Small Factory = TZS 160,000,000

The smallest maximum loss is therefore TZS 110,000,000.


The decision under the Minimax Regret criterion is to choose Medium Factory.

Common questions

Powered by AI

The maximax criterion is an optimistic decision-making approach that focuses on maximizing potential gains by selecting the option with the greatest maximum payoff. It assumes a highly optimistic decision maker who believes that the most favorable outcome will always occur . For example, in the given payoff table, the maximax criterion would lead to the selection of a Large Factory due to its highest potential payoff of TZS 200,000,000 .

Opportunity loss, or regret, measures the cost of not choosing the optimal decision given a particular state of nature. It is calculated as the difference between the actual payoff and the best possible payoff for each scenario. During decision-making under uncertainty, the minimax regret criterion selects an alternative that minimizes the maximum opportunity loss. For instance, in the opportunity loss table, the Medium Factory is chosen because it has the smallest maximum regret of TZS 110,000,000 .

The expected value criterion in decision-making under risk involves calculating the weighted average of all possible payoffs, with weights being their respective probabilities. Decision makers choose the alternative that maximizes this expected value, as it theoretically offers the best long-term payoff. This approach requires known probabilities for uncertain events, and is used when decision makers are risk neutral. The application of this criterion ensures a rational approach to handling risk .

Altering the coefficient of realism (alpha) in the realism (Hurwicz) criterion adjusts the weight between the most optimistic outcome and the most pessimistic outcome, effectively changing the decision maker's position on risk aversion. Increasing alpha shifts the decision towards more optimistic options, while decreasing it reflects a more conservative approach. In the payoff example, varying alpha could switch preference from a Large Factory to a Small Factory if the decision maker's risk aversion increases, demonstrating how subjective risk preferences quantitatively influence objective decision outcomes .

In a certainty environment, decision-making involves choosing alternatives where outcomes are fully known. A decision maker would typically opt for the alternative with the best outcome . Under uncertainty, the decision maker does not know the probabilities of the various outcomes, hence non-probabilistic criteria like maximax, maximin, realism (Hurwicz), equally likely (Laplace), and minimax regret can be applied . In a risk environment, probabilities of outcomes are known, thus probabilistic decision criteria including maximizing expected value and minimizing expected opportunity loss are used .

The maximin criterion considers the most conservative approach by selecting the alternative with the best worst-case scenario, reflecting a pessimistic decision maker. For example, it would opt for Small Factory with a minimum profit of TZS 20,000,000 from the payoff table . Meanwhile, the realism (Hurwicz) criterion combines optimism and pessimism, using a coefficient of realism (alpha) to weigh the best and worst outcomes, thereby balancing risk and optimism. For instance, with a coefficient of 0.6, it favors the Large Factory based on a weighted average of TZS 72,000,000 .

A decision maker would use the minimax regret solution when seeking to minimize potential regret from making suboptimal decisions under uncertainty. This criterion is advisable for those who wish to avoid feeling regret over missed opportunities and can be particularly rational when decision makers want to safeguard against the worst potential regret scenarios. This method results in selecting alternatives based on the smallest possible maximum regret, as shown where the Medium Factory was selected for its lowest maximum opportunity loss of TZS 110,000,000 .

The realism criterion offers a hybrid approach combining both optimistic and pessimistic views through the adjustment of the realism coefficient (alpha), reflecting the decision maker's risk preference. A decision maker can prioritize outcomes by adjusting alpha between 0 (purely pessimistic) and 1 (purely optimistic). As shown, with an alpha of 0.6, the criterion leans towards more optimistic outcomes but still considers worst-case scenarios, aiding balanced decision making by converting subjective risk attitudes into objective choices .

Computing the value of perfect information (VPI) is valuable when weighing the worth of obtaining additional information before making a decision under uncertainty. VPI is applied by comparing the expected outcome with perfect information to the best alternative without it, measuring information's potential to improve decision quality. It is crucial in contexts where decisions involve significant uncertainty and high stakes, providing a quantitative basis for investment in further research or data gathering .

The equally likely (Laplace) criterion assumes all possible states are equally probable, thus calculating an average payoff for each alternative and selecting the one with the highest simple average. This uniformly spreads risk and does not prioritize any specific state of nature over others, unlike strategies that explicitly use probabilities or optimism levels. In the example provided, it results in selecting the Medium Factory due to its highest average payoff of TZS 60,000,000 .

You might also like