sharda_dss10_ppt_02_GE-211565.pptx000000

Business intelligence chapter 2

1. Chapter 2: Foundations and Technologies for Decision Making

2. Learning Objectives Understand the conceptual foundations of decision making Understand Simon’s four phases of decision making: intelligence, design, choice, and implementation Understand the essential definition of decision support systems (DSS) Understand different types of DSS classifications (Continued…)

3. Learning Objectives Learn the capabilities and limitations of DSS in supporting managerial decisions Learn how DSS support for decision making can be provided in practice Understand DSS components and how they integrate

4. Opening Vignette Decision Modeling at HP Using Spreadsheets Background Problem description Proposed solution Results Answer & discuss the case questions...

5. Questions for the Opening Vignette What are some of the key questions to be asked in supporting decision making through DSS? What guidelines can be learned from this vignette about developing DSS? What lessons should be kept in mind for successful model implementation?

6. Characteristics of Decision Making Groupthink Evaluating what-if scenarios Experimentation with a real system! Changes in the decision-making environment may occur continuously Time pressure on the decision maker Analyzing a problem takes time/money Insufficient or too much information

7. Characteristics of Decision Making Decision Support Systems (DSS) Dissecting DSS into its main concepts  Building successful DSS requires a thorough understanding of these concepts

8. Decision Making A process of choosing among two or more alternative courses of action for the purpose of attaining a goal(s) Managerial decision making is synonymous with the entire management process - Simon (1977) Example: Planning What should be done? When? Where? Why? How? By whom?

9. Decision-Making Disciplines Behavioral: anthropology, law, philosophy, political science, psychology, social psychology, and sociology Scientific: computer science, decision analysis, economics, engineering, the hard sciences (e.g., biology, chemistry, physics), management science/operations research, mathematics, and statistics Each discipline has its own set of assumptions and each contributes a unique, valid view of how people make decisions

10. Decision-Making Disciplines Better decisions Tradeoff: accuracy versus speed Fast decision may be detrimental Many areas suffer from fast decisions Effectiveness versus Efficiency Effectiveness  “goodness” “accuracy” Efficiency  “speed” “less resources” A fine balance is what is needed!

11. Decision Style The manner by which decision makers think and react to problems perceive a problem cognitive response values and beliefs When making decisions, people… follow different steps/sequence give different emphasis, time allotment, and priority to each step

12. Decision Style Personality temperament tests are often used to determine decision styles There are many such tests Meyers/Briggs, True Colors (Birkman), Keirsey Temperament Theory, … Various tests measure somewhat different aspects of personality They cannot be equated!

13. Decision Style Decision-making styles Heuristic versus Analytic Autocratic versus Democratic Consultative (with individuals or groups) A successful computerized system should fit the decision style and the decision situation Should be flexible and adaptable to different users (individuals vs. groups)

14. Decision Makers Small organizations Individuals Conflicting objectives Medium-to-large organizations Groups Different styles, backgrounds, expectations Conflicting objectives Consensus is often difficult to reach Help: Computer support, GSS, …

15. Phases of Decision-Making Process Humans consciously or subconsciously follow a systematic decision-making process - Simon (1977) Intelligence Design Choice Implementation (?) Monitoring (a part of intelligence?)

16. Simon’s Decision-Making Process

17. Decision Making: Intelligence Phase Scan the environment, either intermittently or continuously Identify problem situations or opportunities Monitor the results of the implementation Problem is the difference between what people desire (or expect) and what is actually occurring Symptom versus Problem Timely identification of opportunities is as important as identification of problems

18. Decision Making: Intelligence Phase Potential issues in data/information collection and estimation Lack of data Cost of data collection Inaccurate and/or imprecise data Data estimation is often subjective Data may be insecure Key data may be qualitative Data change over time (time-dependence)

19. Application Case 2.1 Making Elevators Go Faster! Background Problem description Proposed solution Results

20. Decision Making: Intelligence Phase Problem Classification Classification of problems according to the degree of structuredness Problem Decomposition Often solving the simpler subproblems may help in solving a complex problem. Information/data can improve the structuredness of a problem situation Problem Ownership Outcome of intelligence phase  A Formal Problem Statement

21. Web and the Decision-Making Process

22. Decision Making: The Design Phase Finding/developing and analyzing possible courses of actions A model of the decision-making problem is constructed, tested, and validated Modeling: conceptualizing a problem and abstracting it into a quantitative and/or qualitative form (i.e., using symbols/variables) Abstraction: making assumptions for simplification Tradeoff (cost/benefit): more or less abstraction Modeling: both an art and a science

23. Decision Making: The Design Phase Selection of a Principle of Choice It is a criterion that describes the acceptability of a solution approach Reflection of decision-making objective(s) In a model, it is the result variable Choosing and validating against High-risk versus low-risk Optimize versus satisfice Criterion is not a constraint! See Technology Insight 2.1

24. Decision Making: The Design Phase Normative models (= optimization) the chosen alternative is demonstrably the best of all possible alternatives Assumptions of rational decision makers Humans are economic beings whose objective is to maximize the attainment of goals For a decision-making situation, all alternative courses of action and consequences are known Decision makers have an order or preference that enables them to rank the desirability of all consequences

25. Decision Making: The Design Phase Heuristic models (= suboptimization) The chosen alternative is the best of only a subset of possible alternatives Often, it is not feasible to optimize realistic (size/complexity) problems Suboptimization may also help relax unrealistic assumptions in models Help reach a good enough solution faster

26. Decision Making: The Design Phase Descriptive models Describe things as they are or as they are believed to be (mathematically based) They do not provide a solution but information that may lead to a solution Simulation - most common descriptive modeling method (mathematical depiction of systems in a computer environment) Allows experimentation with the descriptive model of a system

27. Decision Making: The Design Phase Good Enough, or Satisficing “something less than the best” A form of suboptimization Seeking to achieve a desired level of performance as opposed to the “best” Benefit: time saving Simon’s idea of bounded rationality

28. Decision Making: The Design Phase Developing (Generating) Alternatives In optimization models (such as linear programming), the alternatives may be generated automatically In most MSS situations, however, it is necessary to generate alternatives manually Use of GSS helps generate alternatives Measuring/ranking the outcomes Using the principle of choice

29. Decision Making: The Design Phase Risk Lack of precise knowledge (uncertainty) Risk can be measured with probability Scenario (what-if case) A statement of assumptions about the operating environment (variables) of a particular system at a given time Possible scenarios: best, worst, most likely, average (and custom intervals)

30. Decision Making: The Choice Phase The actual decision and the commitment to follow a certain course of action are made here The boundary between the design and choice is often unclear (partially overlapping phases) Generate alternatives while performing evaluations Includes the search , evaluation , and recommendation of an appropriate solution to the model Solving the model versus solving the problem!

31. Decision Making: The Choice Phase Search approaches Analytic techniques (solving with a formula) Algorithms (step-by-step procedures) Heuristics (rule of thumb) Blind search (truly random search) Additional activities Sensitivity analysis What-if analysis Goal seeking

32. Decision Making: The Implementation Phase “Nothing more difficult to carry out, nor more doubtful of success, nor more dangerous to handle, than to initiate a new order of things.” - The Prince, Machiavelli 1500s Solution to a problem  Change Change management ?.. Implementation: putting a recommended solution to work

33. How Decisions are Supported

34. How Decisions are Supported Support for the Intelligence Phase Enabling continuous scanning of external and internal information sources to identify problems and/or opportunities Resources/technologies: Web; ES, OLAP, data warehousing, data/text/Web mining, EIS/Dashboards, KMS, GSS, GIS,… Business activity monitoring (BAM) Business process management (BPM) Product life-cycle management (PLM)

35. How Decisions are Supported Support for the Design Phase Enabling generating alternative courses of action, determining the criteria for choice Generating alternatives Structured/simple problems: standard and/or special models Unstructured/complex problems: human experts, ES, KMS, brainstorming/GSS, OLAP, data/text mining A good “criteria for choice” is critical!

36. How Decisions are Supported Support for the Choice Phase Enabling selection of the best alternative given a complex constraint structure Use sensitivity analyses, what-if analyses, goal seeking Resources KMS CRM, ERP, and SCM Simulation and other descriptive models

37. How Decisions are Supported Support for the Implementation Phase Enabling implementation/deployment of the selected solution to the system Decision communication, explanation and justification to reduce resistance to change Resources Corporate portals, Web 2.0/Wikis Brainstorming/GSS KMS, ES

38. STRUCTURE OF MATHEMATICAL MODELS FOR DECISION SUPPORT Quantitative models are typ i ca ll y made up of four basic components 1.Result Variables or outcome variables 2. Intermediate Variables 3.Decision Variables 4. Uncontrollable variables In non-quantitative models , the relationships are symbol i c or qualitative . The results of decisions are determined based on th e de c ision made ( i.e. , the va lu es of th e decision variab l es. The modeling process in v olves identifying the variables and relationships among them .

39. Result ( outcome) variables r e flect the level of effectiveness of a system; that is, they indicate how we ll the system performs or attains its goal(s). These variab le s are outputs. Result variables are considered dependent variables . Intermediate Variables I ntermediate result variables are sometimes us e d in modelling to identify intermediate outcomes. Result variables dep e nd on the occurrence of the decision variables and the uncontrollable variab l es.

40. Decision variables describe alternative courses of action. The decision maker contro ls the decision variables. For example , for an investment problem , the amount to invest in bonds i s a decision variab l e. UNCONTROLLABLE VARIABLES, OR PARAMETERS: In any decision-making situation, there are factors that affect th e result variab l es but are not und e r the contro l of th e d ecis ion maker . Either these factors can be fixed, in which case the y are called uncontrollable variables, or parameters, or they ca n vary, in which case th ey are called variables.

41. Intermediate result variables refle c t interm ed iate outcomes in mathematical models. For example , in determining machine schedu lin g, spoi l age is an int e rmediate r es ult variab l e, and total profit i s the r es ult variable (i.e., spoilage i s one determinant of total profit). Another example i s employee sa laries. This co n s titutes a decision variable for management: It determines employee satisfaction ( i.e. , intermediate outcome), which , in turn , determine s th e productivity l evel (i.e., fin a l re s ult).

42. The components of a quantitati ve mod e l are link ed tog e th e r b y math e mati cal (algebraic) expressions - equations or inequalities . A very simp le financial model is P=R-C where P = profit , R = revenue , and C = cost . This eq uation describ es th e r e lationship a mong the variables.

43. CERTAINTY, UNCERTAINTY, AND RISK evaluating and comparing alternatives; during this process, it is necessary to predict the future outcome of each proposed alternative.

44. Decision Making Under Certainty certainty, it is assumed that complete knowledge is available so that the decision maker knows exactly what the outcome of each course of action will be (as in a deterministic environment). The decision maker is viewed as a perfect predictor of the future because it is assumed that there is only one outcome for each alternative

45. Decision Making Under Uncertainty the decision maker considers situations in which several outcomes are possible for each course of action. In contrast to the risk situation, in this case, the decision maker does not know, or cannot estimate, the probability of occun-ence of the possible outcomes.

46. Decision Making Under Risk (Risk Analysis) A decision made under risk (also known as a probabilistic or stochastic decisionmaking situation) Risk analysis (i.e. , calculated risk) is a decision-making method that analyzes the risk (based on assumed known probabilities) associated with different alternatives. Risk analysis can be per formedby calculating the expected value of each alternative and selecting the one with the best expected value .

47. MULTIPLE GOALS, SENSITIVITY ANALYSIS, WHAT-IF ANALYSIS, AND GOAL SEEKING Man age ri a l prob l e ms are seldom evaluated w ith a s in gle s impl e goa l , suc h as profit ma xi mi za ti on. T oday's management sys t ems are mu c h m ore com pl ex, a nd o ne w ith a s in gle goa l is rare.I n stead , man age r s wa n t to a tt a in simultaneous goa ls , so m e of w hi ch m ay co nfli ct. Differen t sta keholder s ha ve d iffer ent goa l s .

48. Certa in difficulties ma y arise w hen a n a l yz in g multiple goals: It is u s uall y difficult t o ob tain a n explicit state ment of the organization ' s goals. The decision maker may c h a n ge th e imp orta n ce assigned to spec ifi c goa l s ove r time or for different decision sce n ar i os. Goals and sub-goals are viewed differently at var i o u s levels of t h e orga ni zatio n a nd w ithin different departme nt s . Goa l s c h a n ge in response to c h a ng es in t h e organization a nd its e n v i ron m en t. Th e r e lati o n s hip between a lt e rn a ti ves a nd their ro l e in deter minin g goa l s may be difficult to quantify. Co mple x problems a r e solved by gro up s of decision maker s , each of w h om has a personal age nd a. • Participants assess the imp ortance (p ri orit i es) of the va ri ous goa l s differently.

49. Sensitivity Analysis Sensitivity analysis attempts to assess the impact of a change in the input data or parameters on the proposed solution.

50. Sensitivity analysis is extremely important in MSS because it allows flexibility and adaptation to changing conditions and to the requirements of different decision-making situations, provides a better understanding of the model and the decision-making situation it attempts to describe, and permits the manager to input data in order to increase the confidence in the model.

51. AUTOMATIC SENSITIVITY ANALYSIS Automatic senstivity analysis is performed in standard quantitative model implementations such as LP. Automatic sensitivity analysis is usually limited to one change at a time, and only for certain variables.

52. TRIAL-AND-ERROR SENSITIVITY ANALYSIS The impact of changes in any variable, or in several variables, can be determined through a simple trial-and-error approach. You change some input data and solve the problem again. When the changes are repeated several times, better and better solutions may be discovered. Such experimentation, which is easy to conduct when using appropriate modeling software, such as Excel, has two approaches: what-if analysis and goal seeking.

53. What-If Analysis What-if analysis is structured as “What will happen to the solution if an input variable, an assumption, or a parameter value is changed? Here are some examples:

54. What will happen to the total inventory cost if the cost of carrying inventories increases by 10 percent? What w ill be the market share if the advertising budget increases by 5 percent? With the appropriate user interface, it is easy for managers to ask a computer model these types of questions and get immediate answers . Furthermore , they can perform multiple cases a nd thereby change the percentage , or any other data in the questio n , as desired. The decision maker does all this directly, without a computer programmer.

55. Goal Seeking Goal seeking calculates the values of the inputs necessary to achieve a desired level of an output (goal). It represents a backward solution approach . The fo ll owing are some examples of goal seeking : What annua l R&D budget is needed for an annual growth rate of 15 percen t by 2018? • How many nurses are needed to reduce the average waiting time of a patient in the emergency room to l ess than 10 minutes?