Using Artificial Intelligence to Enhance Model Analysis
The purpose of mathematical modeling is to generate insights into the decision making environment being modeled. Such insights are often generated through the analysis of several, if not many, related model instances. However, little theory and only a few systems have been developed to support this basic goal of modeling. Nonlinear modeling capabilities of neural networks and related methods can be employed to identify patterns within the multiple ‘what-if’ instances.
The purpose of this paper is to describe a prototype, artificial intelligence-based system, named INSIGHT, which analyzes multiple, related model instances to identify key model parameters and develop insights into how these key parameters interact to influence the model solution.
Unable to display preview. Download preview PDF.
- AbTech, 1990. Abductory Inductive Models—User’s Manual. AbTech, Inc., Charlottesville, VA.Google Scholar
- R. L. Barron, A. N. Mucciardi, F. J. Cook, A. R. Barron and J. N. Craig, 1984. Adaptive Learning in Networks:Development and Applica tions in the U.S. of Algorithms Related to GMDH. In S. J. Farlow (Ed.) Self-Organizing Methods in Modeling:GMDH Type Algorithms. Marcel Dekker, New York, 25–66.Google Scholar
- J.J. Brennan and J.J. Elam, 1986. Understanding and Validating Results in Model-Based Decision Support Systems. Decision Support Systems 2, 49–54.Google Scholar
- Complementary Solutions Inc., 1992. Automate Anytime User Guide. At lanta, GA.Google Scholar
- EXECUCOM, 1992. Interactive Financial Planning System:User’s Man ual. EXECUCOM, Austin, TX.Google Scholar
- S. J. Farlow (Ed.), 1984. Self-Organizing Methods in Modeling:GMDH Type Algorithms. Marcel Dekker, New York.Google Scholar
- W.J. Frawley, G. Piatetsky-Shapiro and C.J. Matheus, 1992. Knowledge Discovery in Databases:An Overview. AI Magazine. Fall, 1992, 57–70.Google Scholar
- A.M. Geoffrion, 1976. The Purpose of Mathematical Programming Is In sight, Not Numbers. Interfaces. 7:1, 81–92.Google Scholar
- H.J. Greenberg, 1990. A Primer of ANALYZE, Working Paper, University of Colorado at Denver.Google Scholar
- F.S. Hillier and G.L. Lieberman, 1990. Introduction to Operations Re search (5e), Holden-Day, Inc. Oakland, CA.Google Scholar
- S.O. Kimbrough, S.A. Moore, C.W. Pritchett and C.A. Sherman, 1992. On DSS Support for Candle-Lighting Analysis. Transactions of DSS-92, 118–135.Google Scholar
- D.W. Kosy and B.P. Wise, 1984. Self-explanatory Financial Planning Mod els. Proceedings of the National Conference of Artificial Intelligence, 176–181.Google Scholar
- D.W. Rosy and B.P. Wise, 1986. Overview of Rome:A Reason-Oriented Modeling Environment in L.F. Psu (ed.), Artificial Intelligence in Eco nomics and Management. Elesvier Science Publishers, North-Holland, 21–30.Google Scholar
- Microsoft, 1992. Microsoft Excel User’s Guide 2 (Version 4.0). Microsoft Corporation.Google Scholar
- M. H. Prager, 1988. Group Method of Data Handling:A New Method for Stock Identification. Transactions of the American Fishery Society 117, 290–296.Google Scholar
- Y.V. Reddy, 1985. The Role of Introspective Simulation in Managerial Decision Making. DSS-85 Transactions, IADSS, University of Texas at Austin, 18–32.Google Scholar
- S. L. Savage, 1992. The ABC’s of Optimization Using What’s Best! LINDO Systems Inc, Chicago.Google Scholar
- A. Saltelli and T. Homma, 1992. Sensitivity Analysis for Model Output. Computational Statistics and Data Analysis 13, 73–94.Google Scholar
- A. Saltelli and M. Marivoet, 1990. Non-parametric Statistics in Sensitivity Analysis for Model Output:A Comparison of Selected Techniques. Relia bility Engineering and Systems Safety 28, 220–253.Google Scholar
- H. M. Wagner, 1993. Global Sensitivity Analysis. To appear.Google Scholar