Download Decision Making under Uncertainty: Maximizing Profits with Decision Theory and more Schemes and Mind Maps Biology of microorganisms in PDF only on Docsity! Decision Theory Prepared by: Jonathan Desenganio Decision Theory represents general approach to decision making CAPACITY PLANNING PRODUCT & SERVICE DESIGN EQUIPMENT SELECTION LOCATION PLANNING DECISION IN OPERATION S CHARACTERISTIC OF DECISIONS UNDER DECISION THEORY Set of possible future condition List of alternatives Pay off for each alternative CAUSES OF POOR DECISION MAKING Bounded Rationality Suboptimization Bounded Rationality The limitations on decision making caused by costs, human abilities, time, technology, and availability of information Suboptimization The result of different departments each attempting to reach a solution that is optimum for that department. STEPS IN DECISION MAKING 1 2 3 4 5 Identify the problem Specify the objectives and criteria for solution Develop suitable alternatives 6 7 Analyze and compare alternatives Select the best alternative Implement the solution Monitor to see that desired result is achieved Decision Environment Certainty Uncertaint y Risk means that relevant parameters such as costs, capacity, and demand have known values means that certain parameters have probabilistic outcomes means that it is impossible to assess the likelihood of various possible future events Certain that the demand will be moderate POSSIBLE FUTURE DEMAND Alternatives Moderate Small Facility 10 Medium Facility 12 Large Facility 2 *Present value in $ millions Identify if profit or cost1 Evaluate the column “Moderate” 2 If profit = select highest If cost = select lowest Identify best payoff3 Select Best Alternative4 Medium Facility Certain that the demand will be high POSSIBLE FUTURE DEMAND Alternatives High Small Facility 10 Medium Facility 12 Large Facility 16 *Present value in $ millions Identify if profit or cost1 Evaluate the column “High” 2 If profit = select highest If cost = select lowest Identify best payoff3 Select Best Alternative4 Large Facility DECISION MAKING UNDER UNCERTAINTY MAXIMAX MAXIMIN LAPLACE MINIMAX REGRET Choose the alternative with the best possible payoff Choose the alternative with the best of the worst possible payoffs. Choose the alternative with the best average payoff of any of the alternatives. Choose the alternative that has the least of the worst regrets. No information is available on how likely the various states of nature are. Maximax Identify if profit or cost1 Evaluate the best payoff for each alternative 2 If profit = select highest If cost = select lowest Identify best out of best payoff3 Select Best Alternative4 Large Facility POSSIBLE FUTURE DEMAND Alternative s Low Moderat e High Small Facility $10* 10 10 Medium Facility 7 12 12 Large Facility (4) 2 16 Best 10 12 16 Laplace Identify if profit or cost1 Calculate the average payoff for each alternative 2 If profit = select highest If cost = select lowest Identify best out of average payoff3 Select Best Alternative4 Medium Facility POSSIBLE FUTURE DEMAND Alternative s Low Moderat e High Small Facility $10* 10 10 Medium Facility 7 12 12 Large Facility (4) 2 16 Average 10 10.33 4.67 Minimax Regret Identify if profit or cost1 Create a regret table2 If profit = select highest If cost = select lowest Evaluate the worst regret for each alternative and identify best payoff 3 Select Best Alternative4 Medium Facility POSSIBLE FUTURE DEMAND Alt Low Mod High Smal l $10* 10 10 Med 7 12 12 Larg e (4) 2 16 a. For each state of nature 10 12 16 b. Subtract the best payoff for each payoff of their respective nature identify the best payoff Regret Table Low Mo d High 0 3 14 2 0 10 6 4 0 Worst 6 4 14 Expected Value Identify if profit or cost1 Calculate the EV for every alternative 2 If profit = select highest If cost = select lowest Identify best EV3 Select Best Alternative4 Medium Facility POSSIBLE FUTURE DEMAND Alt Low Mod High Smal l $10* 10 10 Med 7 12 12 Larg e (4) 2 16 .30 .50 .20 EV 10(.30)+10(.50)+1 0(.20)7(.30)+12(.50)+12( .20)- 4(.30)+2(.50)+16(. 20) EV 10 10.5 3 Exercise New Bridge Build No New Bridge Alternativ e capacity for new store A 1 14 B 2 10 C 4 6 Where A = small, B = medium, and C = Large Assume the payoffs represent profits. Determine the alternative the best alternative 1. If it is certain that there will be: a. a new bridge b. No new bridge 2. If the state of nature is uncertain, determine the best alternative using c. Maximin d. Maximax e. Laplace f. Minimax Regret 3. Determine the best alternative using EV if the probability of there will be no new bridge is 0.3 Solution New Bridge Build No New Bridge Alternativ e capacity for new store A 1 14 B 2 10 C 4 6 Where A = small, B = medium, and C = Large Assume the payoffs represent profits. Determine the alternative the best alternative 1. If it is certain that there will be: a. a new bridge b. No new bridge Solution New Bridge Build No New Bridge Alternativ e capacity for new store A 1 14 B 2 10 C 4 6 Where A = small, B = medium, and C = Large 1. If it is certain that there will be (b) no new bridge New Bridge Build No New Bridge Alternativ e capacity for new store A 1 14 B 2 10 C 4 6 Where A = small, B = medium, and C = Large Small Store Solution New Bridge Build No New Bridge Alternativ e capacity for new store A 1 14 B 2 10 C 4 6 Where A = small, B = medium, and C = Large Assume the payoffs represent profits. Determine the alternative the best alternative 2. If the state of nature is uncertain, determine the best alternative using c. Maximin d. Maximax e. Laplace f. Minimax Regret Solution New Bridge Build No New Bridge Alternativ e capacity for new store A 1 14 B 2 10 C 4 6 Where A = small, B = medium, and C = Large2. If the state of nature is uncertain, determine the best alternative using (c) Maximin New Bridge Build No New Bridge Alternat ive capacit y for new store A 1 14 B 2 10 C 4 6 Where A = small, B = medium, and C = Large Large Store Worst 1 2 4 Solution New Bridge Build No New Bridge Alternativ e capacity for new store A 1 14 B 2 10 C 4 6 Where A = small, B = medium, and C = Large2. If the state of nature is uncertain, determine the best alternative using (e) Minimax Regret New Bridg e Build No New Bridg e Alternati ve capacity for new store A 1 14 B 2 10 C 4 6 Where A = small, B = medium, and C = Large Small Store Regre t Table 3 2 0 0 4 8 Worst 3 4 8 4 14 Solution New Bridge Build No New Bridge Alternativ e capacity for new store A 1 14 B 2 10 C 4 6 Where A = small, B = medium, and C = Large2. If the state of nature is uncertain, determine the best alternative using (e) Minimax Regret New Bridg e Build No New Bridg e Alternati ve capacity for new store A 1 14 B 2 10 C 4 6 Where A = small, B = medium, and C = Large Small Store Regre t Table 3 2 0 0 4 8 Worst 3 4 8 4 14 Solution New Bridge Build No New Bridge Alternativ e capacity for new store A 1 14 B 2 10 C 4 6 Where A = small, B = medium, and C = Large Assume the payoffs represent profits. Determine the alternative the best alternative 3. Determine the best alternative using EV if the probability of there will be no new bridge is 0.3 PRACTICAL APPLICATION OF DECISION TREE Health Care Legal Credit Card Fraud Insurance Risk Analysis New Product or Service Developme nt Location Analysis NODE S Parts of Decision Tree BRANC H SQUARE NODE DECISIO N POINT CIRCULA R NODE CHANC E EVENT Alternativ es Chanc e Event s DIRECTION OF ANALYSIS SQUARE NODE SELECT BEST CIRCULA AR NODE CALCULA TE EV CHOOSE A1 CHOOSE A2 STATE OF NATURE 1 STATE OF NATURE 2 PAYOFF 1 PAYOFF 2 PAYOFF 3 PAYOFF 4 PAYOFF 5 PAYOFF 6 STATE OF NATURE 1 STATE OF NATURE 1 CHOOSE B1 CHOOSE B2 CHOOSE C1 CHOOSE C2 Sample Template of Decision Tree Example Build Small Build Large Low Demand (.4) High Demand (.6) P400 P400 P500 (P100) P700 DO NOTHING OVERTIME DO NOTHING Low Demand (.4) High Demand (.6) P550EXPAN D P500REDU CE PRICE S 55 0 50 490 62 0 5 Build Small Build Large 490 62 0 5 SQUARE NODE SELECT BEST BUILD LARGE THEN REDUCE PRICE Decision Making under RISK EXPECTED VALUE OF PERFECT INFORMATION (EVPI) - How much a decision maker will pay to know with certainty which of state of nature will occur EVPI = Expected Payoff under Certainty - Expected Payoff under Risk EVPI = EPUC - EPUR EPUC Determine the best payoff for each state of nature 1 Multiply the probability to the best payoff of state nature 2 POSSIBLE FUTURE DEMAND Alt Low .30 Moderate .50 High .20 Small $10* 10 10 Med 7 12 12 Large (4) 2 16 1 0 1 2 1 6 10 x .30 =12 x .50 =16 x .20 = 3 6 3. 2 Sum 12. 2 Questions? Suggestions? Comments? THANKYOU