DECISION THEORY

Presented by: Manish Juno Kuldip Juhi Monika Meera Lokesh Loveleena

INTRODUCTION
As the decisions taken by a manager govern the fortunes of business:
Right decisions will have highly profitable effect while the wrong ones may prove to be disastrous.

It is extremely important to choose the appropriate decision.

Decision theory provides a rational approach to the managers in dealing with problems confronted with partial, imperfect or uncertain future conditions.

STEPS IN DECISION THEORY

DECISION-MAKING ENVIRONMENTS
Decisions are made under four types of environments that differ according to the degree of certainty The degree of certainty may vary from complete certainty to complete uncertainty The region that lies in between corresponds to decision-making under risk

Decision-making under conditions of certainty

Decision making under conditions of uncertainty

Decision-making under conditions Of risk

Decision-making under conditions Of conflict

DECISION MAKING UNDER CONDITIONS OF CERTAINTY

Only one state of nature exists

The decision maker simply picks up the best payoff in the table
Though the state of nature is only one, possible alternatives could be numerous

The decision maker enjoys the luxury of having complete information the future

DECISION MAKING UNDER CONDITIONS OF UNCERTAINTY

Under uncertainty criteria, the decision maker has a knowledge about the state of nature that happens but lacks the knowledge about the probability of occurrence. Under conditions of uncertainty, a few decision criteria which could be helpful to the decision makers Choice of the criteria by decision maker is determined by the companys policies & attitude of the decision maker .

1. Maximax criterion or criterion of optimism :

Steps:
a. Select the maximum payoff for each possible alternatives & then choose the maximum payoff

b.

When dealing with the costs, the minimum of each alternative is considered & then the alternative which minimizes the above minimum costs is selected. This is called the Minimin criterion

Alternatives

State of nature high in Rs. moderate in low in Rs. Rs. nil -45000 -80000 -10000

Maximum of row Rs. 50000 70000 <- Maximax 30000

Expand Construct Sub contract

50000 70000 30000

25000 30000 15000

-25000 -40000 -1000

. Maximin criterion or criterion of Pessimism :

Steps:
a. First find the minimum possible payoff for each alternative & then then chooses the alternative with maximum payoff within this group

b. Thus this criterion identifies the worst outcome of each alternative and then selects the best of those worst
c. When dealing with costs, the maximum cost associated with each alternative is considered and the alternative that minimizes the above maximum costs is selected. This is called minimax criterion d. It is also known as WALD criterion

Alternatives high in Rs. Expand Construct Sub contract 50000 70000 30000

State of nature moderate in Rs. 25000 30000 15000 low in Rs. -25000 -40000 -1000 nil -45000 -80000 -10000

Minimum of row Rs. -45000 -80000 -10000 <- Maximin

 The decision maker might regret after making a particular decision. Therefore, the decision maker should try to minimize his regret before selecting a particular strategy or alternative. Regret (often also called opportunity loss) is defined as the difference between the actual payoff and the payoff that would have been obtained if a different course of action had been chosen.

3. MINIMAX REGRET CRITEREON

STEPS:
1. Determine the amount of regret corresponding to each event for every alternative. The regret for jth event corresponding to ith alternative is given by

ith regret = (maximum payoff ith payoff) for the jth event
2. Determine the maximum regret amount for each alternative 3. Choose the alternative which corresponds to the minimum of the above maximum regrets

Alternatives

State of nature moderate high in Rs. low in Rs. in Rs. nil

Minimum of row Rs.

Expand Construct

20,000 0

5000 24000 35000 0 39000 70000 15000 0 0

35000 70000 40000

<-Minimax

Subcontract 40,000

4. HURWICZ CRITERION (CRITERION OF REALISM)

It is a compromise between maximax (optimism) and maximin (pessimistic) decision criteria This concept allows the decision maker to take into account both the maximum and minimum for each alternative and assign them weights according to his degree of optimism (or pessimism). The alternative which maximizes the sum of these weighted payoffs is then selected. Steps: Choose an appropriate degree of optimism , so that (1 - ) represents the degree of pessimism. is called co efficient or index of optimism Determine the maximum as well as the minimum of each alternative and obtain

 If = 0, criterion is too pessimistic; when =1, it is too pessimistic. A value between zero and one may be selected depending upon whether the decision maker leans towards pessimism or optimism. In the absence of strong feeling one way or the other, a value of = 0.5 seems to be a reasonable choice.
State of nature high in Rs. moderate low in Rs. in Rs. nil Maximum of Minimum of row row Rs.

Alternatives

P = .maximum +(1- ). minimum

Expand Construct Subcontract

50000 70000 30000

25000 30000 15000

-25000 -45000 -40000 -80000 -1000 -10000

50000 70000 30000

-45000 -80000 -10000

31000 40000 22000

5. LAPLACE CRITERION OR CRITERION OF RATIONALITY (BAYES CRITERION)

This is based on the principle of insufficient reason. the probabilities associated with the occurrence of various events are unknown, there is not enough information to conclude that these probabilities will be different This criterion assigns equal probabilities to all the events of each alternative decision and selects the alternative associated with the maximum expected payoff Symbolically, if n denotes the number of events and Ps denote the payoff, then the expected value for the strategy, say s 1 is

S1= 1/n [P1 + P2+.Pn]

Alternatives

State of nature high in Rs. moderate low in Rs. in Rs. nil

Expected payoff

Rs.

Expand Construct Subcontract

50000 70000 30000

25000 -25000 -45000 30000 -40000 -80000 15000 -1000 -10000

-1250 -5000 8500

DECISION-MAKING UNDER CONDITIONS OF RISK

More than one states of nature exist

Sufficient information to assign probability A number of decision criteria available to the decision maker

1. EXPECTED VALUE CRITERION

Construct a conditional pay-off table listing the alternative decisions and the various states of nature Calculate the EMV for each decision alternative by multiplying the conditional profits by assigned probabilities and adding the resulting conditional values Select the alternative yielding the highest EMV

2. EXPECTED OPPORTUNITY LOSS CRITERION

Prepare the conditional profit table for each decisionevent combination Determine the conditional opportunity loss Calculate the expected opportunity loss for each decision alternative and adding the values Select the alternative yielding the lowest EOL

3. EXPECTED VALUE OF PERFECT INFORMATION

In the above cases, the occurrence of states of nature was associated with probability. Complete and accurate information about the future demand, referred to as perfect information would remove all uncertainty from the problem. EVPI is an important concept in decision analysis. For a given problem, EVPI represents the maximum amount a person should pay to get additional information on which may be based the decision alternative.

PROBLEM
A dairy firm wants to determine the quantity of butter it should produce to meet the demand. Past records have shown the following demand patterns: Butter costs Rs. 40/kg & is sold at Rs. 50/kg

1. Construct a conditional profit table.

2. Determine the action alternatives with the maximization of expected profit.

Quantity required (kg) 15 20 25

No. of days demand occurred 6 14 20

3. Determine EVPI

30
35 40 50

80
40 30 10

SOLUTION
Conditional Profit table

Possible Demand(kg) Probability(1) 15 20 25 30 35 40 50 0.03 0.07 0.1 0.4 0.2 0.15 0.05 150 150 150 150 150 150 150 -50 200 200 200 200 200 200 -250 0 250 250 250 250 250

Possible stock action (kg) 15 in Rs. 20 in Rs.25 in Rs.30 in Rs.35 in Rs.40 in Rs.50 in Rs. -450 -200 50 300 300 300 300 -650 -150 100 350 350 350 350 -850 -600 -350 -100 150 400 400 -1250 -1000 -750 -500 -250 0 500

 Expected Profit value: EMV=Rs.217.50( From the table)

Possible Demand(kg) Probability(1) 15 20 25 30 35 40 50 EMV 0.03 0.07 0.1 0.4 0.2 0.15 0.05 4.5 10.5 15 60 30 22.5 7.5 150 -1.5 14 20 80 40 30 10 192.5 -7.5 0 25 100 50 37.5 12.5 217.5

Possible stock action (kg) 15 in Rs. 20 in Rs.25 in Rs.30 in Rs.35 in Rs.40 in Rs.50 in Rs. -13.5 -14 5 120 60 45 15 217.5 -19.5 -28 -15 40 70 52.5 17.5 117.5 -25.5 -42 -35 -40 30 60 20 -32.5 -37.5 -70 -75 -200 -50 0 25 -407.5

 Expected Profit Table with Perfect Information

Market Size Expected profit Conditional profit Probability of with perfect under certainty (Rs.) market size information (Rs.)

15 20 25 30 35 40 50

150 200 250 300 350 400 500

0.03 0.07 0.1 0.4 0.2 0.15 0.05 EPPI

4.5 14 25 120 70 60 25 318.5

EVPI= EPPI- max. EMV

DECISION TREE

A visualization of a complex decisionmaking situation in which the possible decisions and their likely outcomes are organized sequentially in the form of a treelike diagram

DECISION TREE

APPLICATION OF DECISION TREE MODEL

Introduction of a New Product

Marketing strategy Make vs. Buy decision

Assets acquisition
Investment decisions

ADVANTAGES
It structures the decision process and helps decision making in an orderly systematic and sequential manner It requires the decision maker to examine all possible outcomes whether desirable or undesirable It communicates decision making process to others in an easy and clear manner

DISADVANTAGES
Decision tree diagrams become more complicated as the number of decision alternatives increases. It becomes highly complicated when interdependant alternatives and dependent variables are present in the problem It analyses the problem in terms of expected values and thus yields an average valued solution

THANK YOU