Constructively learning a near-minimal neural network architecture

Justin Fletcher, Zoran Obradovic

Research output: Contribution to conferencePaper

2 Scopus citations

Abstract

Rather than iteratively manually examining a variety of pre-specified architectures, a constructive learning algorithm dynamically creates a problem-specific neural network architecture. Here we present an revised version of our parallel constructive neural network learning algorithm which constructs such an architecture. The three steps of searching for points on separating hyperplanes, determining separating hyperplanes from separating points and selecting separating hyperplanes generate a near-minimal architecture. As expected, experimental results indicate improved network generalization.

Original languageEnglish (US)
Pages204-208
Number of pages5
DOIs
StatePublished - Jan 1 1994
EventProceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7) - Orlando, FL, USA
Duration: Jun 27 1994Jun 29 1994

Other

OtherProceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7)
CityOrlando, FL, USA
Period6/27/946/29/94

ASJC Scopus subject areas

  • Software

Fingerprint Dive into the research topics of 'Constructively learning a near-minimal neural network architecture'. Together they form a unique fingerprint.

  • Cite this

    Fletcher, J., & Obradovic, Z. (1994). Constructively learning a near-minimal neural network architecture. 204-208. Paper presented at Proceedings of the 1994 IEEE International Conference on Neural Networks. Part 1 (of 7), Orlando, FL, USA, . https://doi.org/10.1109/icnn.1994.374163