Possible low-priced, robust expert systems using neural networks and minimal entropy coding☆

Abstract Within this paper the following research results are presented. A combination of minimal entropy coding and neural network technique is proposed for low-priced and robust expert system building. A “thesis of tractability” describing the size and complexity of ordinary human expertise when using the above-mentioned methods in knowledge engineering is given. Developmental and maintenance costs of expert and decision support systems can be cut down by more than 90 percent below the costs encountered when rule-based shells are used. A PC-based expert system shell (named ZEUS) incorporating the above-mentioned principles has been developed, and two expert systems have been made. Both cases are tractable and significant cost reduction has been verified in specific cases. Principles, techniques, and functionalities of the program ZEUS are described and evaluated.

[1]  Donald F. Specht,et al.  Probabilistic neural networks and the polynomial Adaline as complementary techniques for classification , 1990, IEEE Trans. Neural Networks.

[2]  Jadzia Cendrowska,et al.  PRISM: An Algorithm for Inducing Modular Rules , 1987, Int. J. Man Mach. Stud..

[3]  Daniel Gross.  Induction and ID/3: more powerful than we think , 1988 .

[4]  Nathan Linial,et al.  Results on learnability and the Vapnik-Chervonenkis dimension , 1988, [Proceedings 1988] 29th Annual Symposium on Foundations of Computer Science.

[5]  R. Hecht-Nielsen,et al.  Neurocomputer applications , 1989 .

[6]  Donald Michie,et al.  Machine Learning in the Next Five Years , 1988, EWSL.

[7]  James L. McClelland.  Explorations In Parallel Distributed Processing , 1988 .

[8]  Donald A. Waterman,et al.  A Guide to Expert Systems , 1986 .

[9]  David Haussler,et al.  Learnability and the Vapnik-Chervonenkis dimension , 1989, JACM.

[10]  Stephen I. Gallant,et al.  Connectionist expert systems , 1988, CACM.

[11]  Richard P. Lippmann,et al.  An introduction to computing with neural nets , 1987 .

[12]  J. Ross Quinlan,et al.  Learning Efficient Classification Procedures and Their Application to Chess End Games , 1983 .

[13]  Ken-ichi Funahashi,et al.  On the approximate realization of continuous mappings by neural networks , 1989, Neural Networks.

[14]  David Haussler,et al.  What Size Net Gives Valid Generalization? , 1989, Neural Computation.

[15]  James A. Anderson,et al.  Cognitive and psychological computation with neural models , 1983, IEEE Transactions on Systems, Man, and Cybernetics.

[16]  William A. Thompson,et al.  Finding rules in data , 1986 .

[17]  Vladimir Vapnik,et al.  Chervonenkis: On the uniform convergence of relative frequencies of events to their probabilities , 1971 .