Statistics and decision theory provide methods to help decision makers choose the best course of action from multiple alternatives. Decision making can occur under certainty, risk, or uncertainty. Under certainty, the outcomes of each strategy are known. Under uncertainty, the probabilities of different outcomes are unknown. Under risk, probabilities are known. Different decision criteria exist for each scenario, such as maximizing expected value under risk or choosing the strategy with the minimum potential loss under uncertainty. Statistical decision theory uses concepts from statistics, economics, and psychology to help decision makers optimize outcomes across various decision making environments.
Statistics and decision theory provide methods to help decision makers choose the best course of action from multiple alternatives. Decision making can occur under certainty, risk, or uncertainty. Under certainty, the outcomes of each strategy are known. Under uncertainty, the probabilities of different outcomes are unknown. Under risk, probabilities are known. Different decision criteria exist for each scenario, such as maximizing expected value under risk or choosing the strategy with the minimum potential loss under uncertainty. Statistical decision theory uses concepts from statistics, economics, and psychology to help decision makers optimize outcomes across various decision making environments.
Statistics and decision theory provide methods to help decision makers choose the best course of action from multiple alternatives. Decision making can occur under certainty, risk, or uncertainty. Under certainty, the outcomes of each strategy are known. Under uncertainty, the probabilities of different outcomes are unknown. Under risk, probabilities are known. Different decision criteria exist for each scenario, such as maximizing expected value under risk or choosing the strategy with the minimum potential loss under uncertainty. Statistical decision theory uses concepts from statistics, economics, and psychology to help decision makers optimize outcomes across various decision making environments.
Statistics and decision theory provide methods to help decision makers choose the best course of action from multiple alternatives. Decision making can occur under certainty, risk, or uncertainty. Under certainty, the outcomes of each strategy are known. Under uncertainty, the probabilities of different outcomes are unknown. Under risk, probabilities are known. Different decision criteria exist for each scenario, such as maximizing expected value under risk or choosing the strategy with the minimum potential loss under uncertainty. Statistical decision theory uses concepts from statistics, economics, and psychology to help decision makers optimize outcomes across various decision making environments.
Introduction: Every individual big or small, rich or poor, educated or uneducated, manufacturer or dealer, businessman or housewife, etc has to make decision almost everyday. Without decisions not a single day pass.
Meaning: A set of processor or method which is helpful to decision maker to select wisely the best cause of action from amongstseveral alternatives is called decision theory. When we use synthesis of basic fundamental of statistics, economics and psychology in the decision theory for selecting optimum strategy from the various strategies we called is decision theory as the statistical decision theory. Optimum strategy means that strategy for which the risk or loss is minimum. I. Decision making under certainty. II. Decision making under risk. III. Decision making under uncertainty. IV. Decision making under conflict. V. Decision making under partial information. Scope: The decision theory can be used widely on all the field where we have to take the important decision which is creates. Effects to a variety of a person such as consumers, employees, suppliers, Sweta Salet 1 Statistics for business decision making
distributers, shareholders etc. Hence we can say the decision theory
is applicable in business & commerce, government, finance departments, some political fields and etc. In decision making under certainty the risk of making decisions is too simple. But for that under uncertainty on using certain types of method, rules and laws we can take the best decisions.
Components Or Elements Of Decision
Theory: 1. Acts:
A decision situations arises only when the decision maken more
than ome courses of action open to him. If there is a one alternative there is nothing to decide. For ex: If a student have only one shirt he does not have to decide which shirt to put on. The coloure of action also known as acts or strategies.
2. States of nature:
The selection of any depends upon the external conditions which
are not in the control of decision maker. This States of nature depend upon external factors.
3. Pay-Off:
For a certain act under the ceetain states of a nature we shall we
have the outcome. When this outcome is in rs. otherwise any factor.
4. Pay-Off Matrix:
Sweta Salet 2 Statistics for business decision making
The arisement of different pay-off from different states of nature
and acts known as the pay-off matrix. For ex: x1 x2 x3 x4 x5 x6 x7 x8 x9
5. Regret Table:
When the selection of act we consider the again but sometimes we
may measure loss or cost also. It is known for cost table. Now loss or cost is negative gain in a pay-off table we select a strategy which gives us a maximum pay-off.
Types of Problems in Decision Making
Under Different Environments : Given the pay off matrix for a decision problem, the process of decision making depends upon the situation under which the decision is being made. These situations can be classified into three categories: (a) Decision making under certainty (b) Decision making under uncertainty and (c) Decision making under risk (a) Decision making under certainty : When the result of each strategy can be easily known, the decision maker has to decide the best action that gives maximum pay-off. To understand this matter let us take simple illustrations. Suppose a manufacturer of water glasses can produce and sell a cheaper variety of glasses at Rs. 20 per glass and earns Rs. 2 per glass. He can sell 1000 water glasses per day. On the other hand if Sweta Salet 3 Statistics for business decision making
he produces a superior quality of water glasses and sells at Rs. 30
each, and carns Rs. 4 per water glass. But he can sell only 600 water glasses of superior quality. In this case, we have to decide about the type of water glasses the manufacturer should produce. For cheaper quality of water glasses he can earn 1000 x 2 = 2000 rupees per day. For superior quality of water glasses he can earn 6000 x 4-2400 rupees per day. The profit on the superior quality is more and hence he should produce water glasses of superior quality. (b) Decision making under uncertainty: Under conditions of uncertainty, only pay-offs are known and nothing is known about the likelihood of each state of nature. The problems of decision making under uncertainty can be studied under two heads. 1. Decision making under uncertainty when probabilities of different states of nature are unknown. 2. Decision making under uncertainty when probabilities of different states of nature are known. Different persons have suggested several decision rules for making a decision under such situation. (1) Maximin criterion or Maximin principle: This criterion is used when the decision maker is pessimistic about the future. According to this criterion the minimum pay-off of each act is obtained from the given pay-off matrix. The maximum of these minimum pay-off is identified and the
Sweta Salet 4 Statistics for business decision making
corresponding act is selected. This is explained in the following
In the above, 17 is maximum out of the minimum pay-offs, which
corresponds to act A2. Hence according to maxi-min criterion act A2 should be selected. (2) Maximax decision criterion or Maximax decision principle: This criterion is used when the decision maker is optimistic about future. Maximax implies the maximization of maximum pay-off. The optimistic decision maker locates the maximum pay-off for each possible act. The maximum of these pay-off is identified and the corresponding action is selected. The optimal act in the above example, based on this criterion is A3. (3) Minimax decision criterion or Regret criterion: Minimax criterion attempts to minimize the maximum regret or opportunity loss i.e. mini-max choose the act that is the best of the worst. The practical steps involved in the criterion are as follows: Step 1: Prepare pay-off matrix Sweta Salet 5 Statistics for business decision making
Step 2: Prepare opportunity loss table by subtracting all the Pay-
off elements of an event from the largest pay-off for that event. Step 3: Select the minimum regret for each strategy. Step 4: Select the minimum of these maximum regrets. Step 5: Select the strategy having minimum of maximum regrets. This strategy is based on mini-max criterion. (4) Harwich Decision criterion: It is a more realistic approach which takes into account the degree or index of optimism and pessimism. The practical steps involved in this criterion are as follows: Step 1: Prepare Pay-off Matrix. Step 2: Select the maximum pay-off for each strategy. Step 3: Select the minimum pay-off for each strategy. Step 4: Calculate weighted average for each strategy. Weighted average = α(Maximum pay-off) + (1-α) (Minimum pay-off) Step 5: Select the strategy with the highest weighted average. (5) Laplace Decision criterion: Laplace criterion assumes that all the states of nature are equally likely to occur. Thus, if there are n states of nature, probability of occurrence of each state of nature will be 1/n. Using these equal probabilities, we compute the expected pay-off for each act Xs
Sweta Salet 6 Statistics for business decision making
and the act with maximum expected value is regarded as optimal
and this act is selected as the best act. (c) Decision making under risk: The criterion that is generally accepted as the most appropriate stochastic problems is that of maximizing the expected utility and is often termed either by Expected Monetary Value (EMV) criterion or Expected Opportunity Loss (EOL) criterion. To select the best alternative, we select the alternative for which EMV is highest possible or EOL is the lowest possible. (1) Expected Monetary Value: This is very commonly used technique of decision making. In this technique, probabilities are assigned for different Events (states of nature). The expected pay-off associated with a given combination of an act and event, is obtained by multiplying the pay-off for that act event combination by the probability of occurrence of the given event. The practical steps involved in the calculation of EMV. Step 1: Prepare the pay-off matrix. Step 2: Assign Probabilities to the event. Step 3: Calculate EMV for each act. Step 4: Select the act with the highest EMV. (2) Expected Opportunity Loss: The concept of EOL provides an alternative method of solving a decision problem for the optimal action. Rather than maximizing expected gain, one can minimize the value of regret or the
Sweta Salet 7 Statistics for business decision making
opportunity loss. The steps involved in the calculation of EOL are as
follows: Step 1: Prepare the pay-off table or pay-off matrix. Step 2: Prepare opportunity loss table (Regret table). Step 3: Assign probabilities to the events. Step 4: Calculate EOL for each act. Expected Value of Perfect Information (EVPI): The maximum amount which the decision maker can spend to obtain the perfect information on which to base a given decision is called EVPI. EVPI is difference between expected pay-off with perfect information (EPPI) and Expected maximum pay-off under uncertainty probability of that state of nature. Total of these expected pay-offs gives EPPI. EVPI = EPPI – EMV
