Fast probabilistic self-structuring of generalized single-layer networks

Author

Morris, Robin D. ; Garvin, A. David M

Author_Institution

Inst. Nat. de Recherche en Inf. et Autom., Antipolis, France

Volume

7

Issue

4

fYear

1996

fDate

7/1/1996 12:00:00 AM

Firstpage

881

Lastpage

888

Abstract

An algorithm is presented for determining the subset of the basis functions of a generalized single-layer network (GSLN) needed to solve the classification problem defined by the training data. A Markov chain Monte Carlo sampling technique is used to traverse the space of models having a low sum squared error (SSE). The frequency of a term´s inclusion is an indication of its importance to the classifier. Fast, iterative updates can be used for the matrix calculations needed. Theoretical results for the required length of the chain needed to obtain good discrimination between functions fitting the data and those modeling the added noise are given, and these are confirmed by experiment

Keywords

Markov processes; Monte Carlo methods; matrix algebra; pattern classification; self-organising feature maps; Markov chain Monte Carlo sampling technique; basis functions; classification problem; fast probabilistic self-structuring; generalized single-layer networks; iterative updates; sum squared error; Cost function; Frequency; Helium; Laboratories; Monte Carlo methods; Neural networks; Polynomials; Signal processing; Signal processing algorithms; Training data;

fLanguage

English

Journal_Title

Neural Networks, IEEE Transactions on

Publisher

ieee

ISSN

1045-9227

Type

jour

DOI

10.1109/72.508931

Filename

508931