Vous êtes sur la page 1sur 7

Lecture 1 INTRODUCTION TO DECISION SUPPORT SYSTEMS

A. HISTORY [Power, 1999]

Management Decision Systems:


Computer-Based Support for Decision Making [Scott-Morton (1971) Ph.D thesis at MIT]
How computers and analytical models help or support managers to make key decisions Experiment in which managers actually used a Management Decision System (MDS)
l

Pioneers in the DSS field


Theoretical studies on organizational decision-making by Simon and Newell at Carnegie Institute of Technology in 60s Research on interactive computing at MIT & man- machine interfaces [Gerrity, 1970]

used by marketing and production managers to coordinate production planning for laundry equipment

Pioneered implementation, definition and research testing of a specific DSS.

Decision Calculus

[John Little,1970]

Management Information Systems:

Expanded frontiers of computer-supported modeling used time-shared computers Identified criteria for designing models to support management decision making Criteria included: robustness, ease of control, simplicity, and completeness of relevant detail.

Decision Support Systems: An


Organizational Perspective [Keen and Scott-Morton, 1978]
Provided comprehensive behavioral orientation to DSS analysis, design, implementation, evaluation and development

Conceptual Foundations, Structure, and Development [Gordon Davis (1974) Univ. of Minn.]
Created broad foundation for DSS research and practiceintroduced concept of MIS

Foundations of Decision Support Systems


Broadened the field of DSS and provided an abstract framework for designing a DSS

[Bonczek, Holsapple, and Whinston (1981)]

Building Effective Decision Support

Systems [Sprague and Carlson (1982)]


Provided a practical, understandable overview of how organizations could and should build a DSS DSS with an enterprise-wide scope

In the 1980s and 1990sgrowth of


Discipline-Oriented DSSs

Concepts of Data Warehousing--1990 On-Line Analytical Processing (OLAP)


Comparable to data warehouses, but analytical tools more extensive Converged with concepts of Executive Information Systems (EIS) 5

Advanced applications of:


Engineering Decision Support Systems Spatial Decision Support Systems (with GIS) Artificial intelligence Human-computer interaction Software engineering Telecommuncations Information sciences Operations research and management science

B. APPROACHES TO DECISION-MAKING

Descriptive approach

Related to MIS for structured or programmed decisions Behavioral--how decision makers actually behave
Human intuition and judgment Qualitative analysis

Normative approach

Prescribing optimal behavior or how decisions should be made Rational approach; quantitative Operations research, mathematical programming and management science How to choose from set of alternatives Little focus on framing actual problems, setting goals or implementation issues
7

Hybrid approaches

Extensive use of heuristics in decision-making Application of knowledge engineering and expert systems to decision support Applicable to unstructured or fuzzy problems Combining normative and descriptive approaches

C. HEURISTIC PROBLEM SOLVING

Human expertise:
"An expert is a person who, because of training and experience, is able to do things the rest of us cannot; experts are not only proficient but also smooth and efficient in the actions they take. Experts know a great many things and have tricks and caveats for applying what they know to problems and tasks; they are also good at plowing through irrelevant information in order to get at basic issues, and they are good at recognizing problems they face as instances of types with which they are familiar." Johnson (1983):

Problem solving
Initial situation goal) Desired situation (or

Human problem solving is heuristic


Procedures that are efficient and work most of time, though at times, imperfectly Rules of thumb

10

Characteristics of human experts


Performance [reach solutions with incomplete and uncertain knowledge--use caveats, or default reasoning] Discrimination [...of what is relevant
l

Dealing with risk and uncertainty


Normative approaches utilize probability theory to deal with uncertainty
l

common sense]

E.g., use Bayes' Theorem: compute how probabilities change as events are perceived

Pattern recognition [perceive recurring patterns] Domain knowledge


l

requires training and experience

Unconscious knowledge [cannot explain why make certain decisions --most often correct] Forward inferencing Theoretical and experiential knowledge

Evidence shows that human decision makers do not really employ a Bayesian approach; use simple heuristics They look at recurring patterns, not probabilities Familiar recall often dictates decisions
l

Anchoring"--or bias towards initial values

11

12

Do decision makers attempt to


maximize expected utility?
tend to be more biased towards minimizing risk DMs tend to concentrate on a single important attribute, and then bring others in as needed E.g., focus on cost first when buying car-then consider make, style, color, etc.

"satisficing" choice behavior--look for


decision that is "good enough"--rather than optimal decision makers often deal with unstructured problems by attempting to decompose them into more familiar, structured subproblems

13

14

D. IMPORTANCE OF NORMATIVE DECISION MAKING

Rational framework Transitive


A--->B and B---->C then A---->C X better than Y and Y better than Z; if

DMs have difficulties grasping normative methods requiring high levels of cognition:
too abstract and complex lack of easy -to-use software do not match real problems or the "way I do it"

Normative approaches may speed up learning processes--better decision making Decision making processes vs. outcome of decision
good good good bad bad bad good bad 15

Consistency important

conclude however that X not better than Z, inconsistent or irrational]

If we agree on set of rules, then

decisions should be made to be consistent with them


16

Operations research (OR)


excellent for repetitive decision processes --less used for strategic, high risk decision making more focus on efficient problem solving process rather than the efficacy of the resulting decision failure to involve decision makers failure to be sensitive to human and institutional contexts in decision making failure to consider ends as well as means excellent for structured problems better solutions for certain specific classes of problems 17 18

Emergence of multicriteria decision


analysis
Watson and Buede (1987): Conflicts between objectives Uncertainty on outcomes Value preferences made between objectives under complete certainty Risk aversion

Failures in OR

Utility theory

Successes in OR

E. WHAT IS A DSS?

Klein and Methlie (1995): "A computer


information system that provides information in a given domain of application by means of analytical decision models and access to databases, in order to support a decision maker in making decisions effectively in complex and ill-structured tasks."
19

F. COMPONENTS OF A DSS
1. Data base management subsystem [DBMS]
Coordination, integration, integrity, storage and extraction of information Separation of data and decision models

2. Display (dialog); user/model interaction; alerting


Manipulate model--check logic Input data during model execution Define display preferences; colors, windowing

20

3. Problem analysis and modeling


Statistical data analysis Forecasting algorithms Simulation, optimization and other OR

Decision Maker

Dialog Subsystem

methods
Knowledge encoding
Static Database

Central Processing Unit Planning Models Data Base Management System Model Base Management System Management Models Operations/ Control Models

All DSSs must contain all 3 of these components to some extent!!


21

Dynamic Database

Data Subsystem

Models Subsystem

22

G. CHARACTERISTICS OF ENGINEERING DSS

Question

Language Processing

Information Collection

Stored Info.

Preferences, judgments, intuition, experience of decision maker essentialend user focus Provides variety of solution needs:
Information Modeling Data centered Flexible Based on user preferences Computation Data manipulation

Decision Maker

H. FUNCTIONS OF A DSS
Problem Recognition

Solution procedures should be


- Allow for numerous and conflicting criteria - Solutions must be found in limited time 23 Analysis Model Formulation

24

The Decision Pyramid


Decision Judgment Knowledge Behavior Models Relationships Data
D. Ullman, 2 0 0 1

Application to Systems Structure of a System


Collection of objects such technology, facilities, resources, concepts, and procedures to perform an identifiable function or serve a goal Inputs Processes Outputs Feedback Environment and boundaries (closed or open)

Performance measures
25

Effectiveness [doing the right thing] Efficiency [doing the thing right] 26

Models
Simplified representation or abstraction of a real system for analysis of a specific problem Problem focus dissatisfaction with the way things are going; difference between desires and reality Cost of modeling lower Compression of time Manipulation (changing variables) much easier Cost of making mistakes lower Able to calculate risks Analysis of large number of solutions 27

Model formulation
Raw information to usable information Related to scientific method
l l

Benefits

Observation of problem Formulation of logical, mathematical model to explain system


defining variables (independent and dependent) defining causal relationships

l l

Prediction of problem solution Verification of solution

28

Model formalism

INPUTS Uncontrollable

MODEL Analytical equations Rules Constraints Procedures

OUTPUTS

Model validation Model manipulation

Analytical or heuristic Physically -based or conceptual Lumped or distributed Deterministic or Stochastic Discrete or continuous Dynamic or static
Computation Sensitivity analysis Impact and tradeoff analysis
l

Decision Variables

Criteria or Goal Attainment

Simulation or Descriptive Model answering what if questions

Multi-criteria decision making

29

30

INPUTS Uncontrollable

MODEL Analytical equations Rules Constraints Procedures

OUTPUTS

I. REQUIREMENTS FOR DSS SOFTWARE

Criteria or Goal Attainment

End user friendly and interactive; self contained; on-line; End user designed Easy access to pertinent information
Data access and meaning Variables; use of scripting languages (e.g., PERL) Decision analysis techniques Statistical, simulation and optimization tools Syntax and semantics necessary for usage Knowledge about rules How information displayed 32

Decision Optimal or Heuristic Variables Search Algorithm

Optimization or Prescriptive Model prescribing solutions

31

High interaction between user and system


"...while quantitative methods and normative models are developing rapidlynevertheless intuition, judgment and experience remain essential factors in the process of exploration of alternatives and the search for adequate tools [Klein and Tixier (1971)]

J. VALUE OF DSS
lIncreased quality of decision making
Problems identified more easily Rapid assimilation through graphical display Generation of more alternatives Compare alternative more easily

Ability to evolve to user needs Portability Reliability

lImproved efficiency
Cost reduction Delay reduction Clear documentation
33 34

Acceptable performance

lBetter communication among decision makers lBetter documentation of decision making lImproved learning -- simulators as DSS

K. POTENTIAL DRAWBACKS OF DSS [Winograd and Flores, 1986]


l l l l l l

Depersonalization of decision making False belief in objectivity Assumption of relevance Obscuring responsibility Unanticipated impacts Unintended transfer of power

35

36

BIBLIOGRAPHY
Bonczek R. H., C. W. Holsapple, and A. Whinston. Foundations of , Decision Support Systems . Academic Press, 1981. Eom, Sean B., "Assessing the Contributions of Systems Science to the Development of the Decision Support System Research Subspecialties". AIS Americas Conference, Phoenix, AZ, August 1996. Gerrity, T. P., Jr. "The Design of Man-Machine Decision Systems". Sloan Management Review , Vol. 12, No. 2, pp. 59-75, Winter 1971. Holsapple, C. and A. Whinston (eds.), Decision Support Systems: Theory and Application, Springer -Verlag , New York, 1987 House, W. C., Decision Support Systems , McGraw-Hill, New York, 1991 Johnson, P. E., What Kind of an Expert Should a System Be?, The Journal of Medicine and Philosophy, Vol. 8, pp. 77-07, 1983.

Keen, Peter G. W. and Michael S. Scott Morton. Decision Support Systems: An Organizational Perspective. Reading, MA: Addison-Wesley, Inc., 1978 ISBN 0-201-03667-3. Keeney, R. and H. Raiffa, Decisions with Multiple Objectives, Preferences, and Value Tradeoffs, John Wiley & Sons, New York, 1976. Klein, M. and L. B. Methlie, Knowledge-based Decision Support Systems with Applications in Business, John Wiley & Sons Chichester, UK, 1995. Labadie, J., "Decision Support Systems in Water Resources," in Stochastic Hydrology and Its Uses in Water Resource Systems , J. Marco, R. Harboe, and J. Salas, eds., Kluwer Academic Publishers, Boston, 1993. Labadie, J.and C. Sullivan, "Computerized Decision Support Systems for Water Managers," Journal of Water Resources Planning and Management, ASCE, Vol 112, No. 3, pp. 299307, July 1986.

37

38

Little, J. D. C. "Models and Managers: The Concept of a Decision Calculus". Management Science, Vol. 16, No. 8, pp. B466-485, April 1970. Newell, A. and H. A. Simon, Human Problem Solving, PrenticeHall, Englewood Cliffs, New Jersey, 1972. Olson, D. and J. Courtney, Decision Support Models and Expert Systems , Dame Publications, Inc., 1997. Power, D.J. A Brief History of Decision Support Systems. DSSResources.COM, World Wide Web, http://DSSResources.COM/history/dsshistory.html, 1999. Rockart, J. F. "Chief Executives Define Their Own Data Needs," Harvard Business Review , Vol. 67, No. 2 (March-April 1979), pp 81-93. Scott Morton, Michael S. Management decision systems; computer-based support for decision making. Boston, Division of Research, Graduate School of Business Administration, Harvard University, 1971.

Sprague, R. H. and E. D. Carlson. Building Effective Decision Support Systems . Englewood Cliffs, N.J., Prentice-Hall, Inc., 1982 Sage, A. P., Decision Support Systems Engineering , John Wiley & Sons, Inc., New York, 1991. Turban, E., Decision Support and Expert Systems: Management Support Systems, Prentice-Hall, Englewood Cliffs, New Jersey, 1995. Ullman, D. The Ideal Engineering Decision Support System, Robust Decisions, Inc., Corvallis, OR, 2001. Watson, S. R. and D. M. Buede, Decision Synthesis: The Principles and Practice of Decision Analysis, Cambridge University Press, Cambridge, Mass. Winograd, T. and F. Flores, Understanding Computers and Cognition, AblexPress, Norwood, NJ, 1986.

39

40

Vous aimerez peut-être aussi