A basis function approach to programming concurrent voting systems to perform selection tasks

Voting is one method available to knowledge-based systems for dealing with selection uncertainty. Evidence on the validity of one or more selections is collected. Each piece of evidence then casts a ''vote'' for the defended selection. In this way, evidence in favor of various alternatives is accumulated. The numeric value of each vote depends on the applicability of the associated evidence. Thus, applicability controls the contribution each piece of evidence makes toward any specific selection. One difficult problem in any voting mechanism is finding a way to specify the numeric value of each vote in a given circumstance. One method of developing this situationally-dependent specification is the use of Gaussian radial basis functions by a network of processors operating concurrently. The technique discussed in this paper takes neural network technology as an inspiration for the implementation of such an information processing scheme. It provides a numerical method for solving a problem faced by symbolic knowledge-based systems, that of examining and deciding between alternatives using a large body of evidence. Due to the concurrent nature of the implementation, there is a great potential for real-time execution. The programs for the network of processors are developed in O(N) or O(N^2) time, depending on how the programs are developed. (N refers to the amount of evidence.) The network itself, once programmed, executes in O(1) time, if sufficient numbers of processors are employed. The number of processors modelled by a fully programmed network is O(N). This bound on execution and programming resources is predictable during problem specification. These advantages favor logistical supportability, a necessary system implementation requirement if advanced computing methods are to be incorporated into fielded equipment. This paper offers both theory and implementation. An illustrative example is given as well as references to related papers.

[1]  Amnon Yariv,et al.  The CCD neural processor: a neural network integrated circuit with 65536 programmable analog synapses , 1990 .

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

[3]  D. I. Blockley,et al.  Uncertain Inference in Knowledge‐Based Systems , 1987 .

[4]  Donald F. Specht,et al.  A general regression neural network , 1991, IEEE Trans. Neural Networks.

[5]  Larry R. Smeltzer,et al.  Management: The Competitive Edge , 1993 .

[6]  L H Meskin,et al.  The competitive edge. , 1994, Journal of the American Dental Association.

[7]  Daniel J. Power,et al.  Soelberg's Job Search and Choice Model: A Clarification, Review, and Critique , 1985 .

[8]  Bruce W. Suter,et al.  The multilayer perceptron as an approximation to a Bayes optimal discriminant function , 1990, IEEE Trans. Neural Networks.

[9]  Steven C. Gustafson,et al.  Stretch and Hammer Neural Networks for N-Dimensional Data Generalization , 1992 .

[10]  Pietro Burrascano,et al.  Learning vector quantization for the probabilistic neural network , 1991, IEEE Trans. Neural Networks.

[11]  D. R. Collins,et al.  Considerations for neural network hardware implementations , 1989, IEEE International Symposium on Circuits and Systems,.

[12]  William B. Gevarter.  The basic principles of expert systems , 1991 .

[13]  R. Sanner,et al.  Gaussian Networks for Direct Adaptive Control , 1991 .

[14]  Julius T. Tou,et al.  Pattern Recognition Principles , 1974 .

[15]  K. Ramachandran,et al.  Mathematical Statistics with Applications. , 1992 .

[16]  Daniel Kersten,et al.  Introduction to neural networks , 1993 .

[17]  Glenn Shafer,et al.  A Mathematical Theory of Evidence , 2020, A Mathematical Theory of Evidence.

[18]  P. Koch,et al.  Digital implementation of a neural network , 1989, International 1989 Joint Conference on Neural Networks.

[19]  Jack S. Walicki,et al.  Inexact Reasoning In Expert Systems , 1987, Other Conferences.

[20]  Daniel R Zahirniak.  Characterization of Radar Signals Using Neural Networks , 1990 .

[21]  Keith L. Clark,et al.  Negation as Failure , 1987, Logic and Data Bases.

[22]  Donald F. Specht,et al.  Probabilistic neural networks , 1990, Neural Networks.