Focus: Semi-structured decisions Analytic v. Heuristic Decisions in the Problem Solving Phases Multiple-Criteria Methods
Technology has always been used to augment the human decision-making process Now, however, it is being used to replace the human decision-making process The more we automate, the less we make the decisions that affect us Yet we seem willing, eager, to give up that responsibility without considering consequences, or (then) resisting when they manifest
Intelligence: Using the information you have to make good decisions Bad decisions can have ramifications, haunt you for years to come Business Intelligence Using the information you have to make good business decisions Bad decisions will negatively impact your company in significant ways for a long time
Decision Support Systems Models (many types) Descriptive Static / Dynamic Mental Decision trees / tables Simulations Flow charts / Diagrams
Decision Support Systems Organize the information used in the decision making process Support, not replace Can affect or change the users decision making process Best for supporting complex, unique decisions Highly individualized Focus is on the process, not output
Making fast-paced decisions More decisions are being made in less time Missed opportunities 75% of workers, 80% of managers Decisions not made fast enough Don t consider all information Can t consider all information in time Too many sources
Decision Making Under Risk Certainty Uncertainty Risk
Decision Making Styles Analytic Heuristic Implications for DSS development
Problem-Solving Phases Intelligence Design Choice Why and where do bottlenecks happen?
Dimensions of Semi-Structured Decisions Degree of decision-making skill required The degree of problem complexity Number of criteria considered
Semi-Structured Decisions in the Problem-Solving Phases Intelligence phase decision support systems Identify the problem Define the problem Assign a priority to the problem
Semi-Structured Decisions in the Problem-Solving Phases Design phase decision support systems Identify alternatives Quantify alternatives Establish performance criteria May help decision maker identify appropriate alternatives
Semi-Structured Decisions in the Problem-Solving Phases Choice phase decision support systems Choice always left up to decision-maker Present methods of selection Help organize and display the information
Multiple-Criteria Decision Making The Trade-Off Process Two columns Pros Cons Updated approach using positives only
The Trade-Off Approach A B C D Eat at McDonald s Eat at health food store Tastes Better 1 Healthier 1 Can get fries 1 Wider selection 1 Have happy meals 1 Common food alts 1 Many locations 1 Discover new tastes 1 Nifty mascot 1 Feel Better 1 Cheaper 1 6 5
Multiple-Criteria Decision Making Weighting Methods Each attribute is assigned a weight (value) Each alternative is assigned a grade for each attribute Choose alternative with highest total score
Decision Making Weighting Methods Weight Attributes Car 1 Car 2 Car 3 0.2 Manual transmission 2 2 1 0.15 Power seats 4 4 4 0.1 Engine power 5 5 10 0.15 Convertible 10 5 6 0.2 Media player inputs 10 10 10 0.05 Gas mileage 8 10 4 0.1 Color 8 6 3 0.05 Trunk space 6 7 5 Total 6.5 5.7 5.45
Multiple-Criteria Decision Making Sequential Elimination by Lexicography Attributes are ranked according to importance Each alternative is assigned a grade for each attribute Only choose alternatives with highest score Repeat until one choice is left
Sequential Elimination by Lexicography Attributes Car 1 Car 2 Car 3 1 Manual transmission 10 10 10 2 Power seats 2 2 1 3 Engine power 5 5 10 4 Convertible 4 4 4 5 Media player inputs 8 10 4 6 Gas mileage 10 5 6 7 Color 8 6 3 8 Trunk space 6 7 5
Sequential Elimination by Lexicography Attributes Car 1 Car 2 Car 3 1 Manual transmission 10 10 10 2 Power seats 2 2 1 3 Engine power 5 5 10 4 Convertible 4 4 4 5 Media player inputs 8 10 4 6 Gas mileage 10 5 6 7 Color 8 6 3 8 Trunk space 6 7 5
Sequential Elimination by Lexicography Attributes Car 1 Car 2 Car 3 1 Manual transmission 10 10 10 2 Power seats 2 2 1 3 Engine power 5 5 10 4 Convertible 4 4 4 5 Media player inputs 8 10 4 6 Gas mileage 10 5 6 7 Color 8 6 3 8 Trunk space 6 7 5
Multiple-Criteria Decision Making Sequential Elimination by Conjunctive Constraints Set up constraints Each alternative must satisfy constraints If not, that alternative is eliminated
Sequential Elimination by Conjunctive Constraints Attributes Constraints Car 1 Car 2 Car 3 Manual transmission 5spd/6spd 5 spd T 4 spd F 6 spd T Power seats Yes Yes T Yes T Yes T Engine power >250hp 400 T 220 F 350 T Convertible Yes No F Yes T Yes T Fuel tank >15 gal 15 T 12 F 18 T Gas mileage >15mpg 12 F 20 T 16 T Color Red/blue/black Black T Red T Green F Trunk space >12 cu. ft. 14 T 8 F 12 T F F F