Decision Theory

Team Members
Shweta Mishra Rejoy K.J.
Ranjeet Kumar Ronak Bhadani
Ravi Kant Ravi Kumar
 What is Decision theory

• Decision theory in mathematics and statistics is concerned

with identifying the values, uncertainties and other issues
relevant in a given Decision, its rationality, and the resulting
optimal Decision.

• Decision theory provides a rational approach to the

managers in dealing with problems confronted with partial,
imperfect or uncertain future conditions.
 Decision Process : Steps involved

• List all the viable alternatives

• Identify the State of nature.
States of nature refer to a set of possible future conditions, or events, beyond
the control of the decision maker that will be the primary determinants of the
eventual consequence of the decision.
• Construct a payoff table
The consequence resulting from a specific combination of a decision
alternative and a state of nature is a payoff.
• Select optimum Decision criterion
 Decision Making environment
• Condition of certainty
only one state of nature exists for each alternative.

• Condition of uncertainty
More than one state of nature exists but the lack of sufficient knowledge.

• Condition of Risk
More than one state of nature exists but has the sufficient knowledge.

• Condition of conflict
Two or more opponents with conflicting objectives try to make decisions.
 Condition of certainty

• Only one state of nature exists.

• The particular state of nature is associated with probability 1.

Alternatives State of nature

(High Product Demand)

Expand Rs. 50,000

Construct Rs. 70,000

Sub construct Rs. 30,000

 Condition of certainty

• Only one state of nature exists.

• The particular state of nature is associated with probability 1.

Alternatives State of nature

(High Product Demand)

Expand - Rs. 25,000

Construct - Rs. 40,000

Sub construct - Rs. 1,000

 Condition of uncertainty
• Criterion of optimism:
The decision maker uses optimistic approach. i.e. in the case of
Product demand, Profit – maximization (Maximax)
Cost – Minimization (Minimin)
To find the maximum possible payoff for each possible alternative with
maximum payoff within this group.

Alternatives State of nature (Product Demand) Maximum of

Row
High Moderate Low Nil
(Rs.) (Rs.) (Rs.) (Rs.)
Expand 50,000 25,000 -25,000 -45,000 50,000
Construct 70,000 30,000 -40,000 -80,000 70,000
(Maximax)
Sub construct 30,000 15,000 -1,000 -10,000 30,000
 Condition of uncertainty
• Criterion of Pessimism (wald criterion):
The decision maker uses Pessimistic approach. He wants to
Product demand, Profit – Maximizing the minimum (Maximin)
Cost – Minimizing the maximum (Minimax)
To find the maximum (minimum) possible payoff for each possible
alternative with minimum (maximum) payoff within this group.

Alternatives State of nature (Product Demand) Maximum of

Row
High Moderate Low Nil (Rs.)
(Rs.) (Rs.) (Rs.)
Expand 50,000 25,000 -25,000 -45,000 -45,000
Construct 70,000 30,000 -40,000 -80,000 -80,000
Sub construct 30,000 15,000 -1,000 -10,000 -10,000
(Maximin)
 Condition of uncertainty
• Minimax Regret criterion (Savage criterion):
The decision maker attempt to minimize regret before actually selecting a
particular strategy. Steps involve in regret method is :
1. Determine the amount of regret corresponding to each event for every alternative.
ith regret = (maximum payoff – ith payoff) for the jth event.
2. Determine the maximum regret amount for each alternative.
3. Choose the alternative which is the minimum of the above maximum.

Alternatives State of nature (Product Demand) Maximum of

Row
High (Rs.) Moderate (Rs.) Low (Rs.) Nil (Rs.)

Expand 20,000 5,000 24,000 35,000 35,000

Construct 0 0 39,000 70,000 70,000
Sub construct 40,000 15,000 0 0 40,000
 Condition of uncertainty
• Hurwicz Criterion (criterion of Realism)
It is also called weighted average criterion. It is a compromise
between the Maximax (optimism) and Maximin (Pessimism). It uses the
concept of coefficient of optimism (α). Steps followed in this criterion:
1. Choose an appropriate degree of optimism α, so that (1-α) is called the
degree of pessimism.
2. Determine the maximum and minimum of each alternative and obtain P
= α. Maximum + (1-α). Minimum.
3. Choose the alternative that yield the maximum value of P.
Alternatives State of nature (Product Demand) Maximum Minimum Value
of Row of Row Of P
High (Rs.) Moderate (Rs.) Low (Rs.) Nil (Rs.)

Expand 50,000 25,000 -25,000 -45,000 50,000 -45,000 31,000

Construct 70,000 30,000 -40,000 -80,000 70,000 -80,000 40,000
Sub construct 30,000 15,000 -1,000 -10,000 30,000 -10,000 22,000
 Condition of uncertainty
• Laplace Criterion (criterion of Rationality):
Since the probabilities associated with the occurrence of various
events are unknown, there is not enough information to conclude that
these probabilities will be different. That’s why we assigns probabilities to
all the events of each alternative decision and selects the alternative
associated with the maximum expected payoff.
E = 1/n [P1 + P2 + P3 + …….. + Pn ]

Where E is the expected payoff and

N is the number of events
P’s is the payoff.
 Condition of Risk
• More than one state of nature exist
• Sufficient information
Under the condition of Risk, a number of decision criteria are available which could
be of help to the decision-maker:
Expected value criterion
This criterion requires the calculation of the expected value of each decision
alternative which is the sum of the weighted payoff’s for that alternative.
The weights are the probabilities assign to the state of nature.
Steps involved :
1. Construct a conditional pay off table listing the alternative decision and the
various state of nature.
2. Calculate the EMV for each decision alternative by multiplying the conditional
profits by assigned probabilities and adding the resulting conditional values.
3. Select the alternative with the highest EMV.
 Decision Tree
• What do you understand by decision tree analysis?
• A decision tree is a graphical representation of the decision
process indicating decision alternatives, state of nature,
probabilities attached to the state of nature and conditional
benefits and losses.

• What is node in a decision Tree?

• A decision node is usually represented by a square and
indicates places where a decision maker must select one
alternate cause of action among the available action.
 Decision Tree
• What is backward pass?
• The calculation of late finish times (dates) for all uncompleted network activities for a
specific project by subtracting durations of uncompleted activities from the scheduled
finish time of the final activity.

• What are the basic steps involve in constructing of a decision tree?

• Steps involve in decision tree analysis:
1. Identify the decision points and the alternative courses of action at each decision point
systematically.
2. Determine the probability and the payoff associated with each course of action.
3. Commencing from the extreme right end, compute the expected EMV.
4. Choose the course of action that yields the best payoff for each of the decisions.
5. Proceed backwaeds to the next stage of decision points.
6. Repeat above steps till the first decision point is reached.
7. Finally, identify the courses o faction to be adopted from the beginning to the end.
 Decision Tree
• Advantages:
1. It structures the decision process and helps decision making in and orderly,
systematic and sequential manner.
2. Examine all possible outcomes.
3. Easy and clear to communicate.
4. Consider the time sequence.

• Limitations:
1. Becomes more complicated as the number of decision alternatives
increases.
2. It assumes that utility of money is linear with money.
3. It yields an average valued solution.
4. Inconsistency in assigning probabilities for different events.