PROBLEM OF INTERNAL REPRESENTATIONS IN NEURAL NETS: CONCEPTS, IMPLICATIONS, AND A PROPOSED METRIC.

Stephen Helms Tillery, Nathan Combs

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Scopus citations

Abstract

A conceptual tool with which to assess the learning ability of neural networks is proposed. What is desired is a general-purpose tool with which to appraise the representations of knowledge of a system that are formed over time and experience. The authors suggest that such a tool should not be limited to any one neural simulation paradigm. The requirements of such an approach are examined, and a metric for this purpose is proposed.

Original languageEnglish (US)
Title of host publicationUnknown Host Publication Title
EditorsMaureen Caudill, Charles T. Butler, San Diego Adaptics
Place of PublicationSan Diego, CA, USA
PublisherSOS Printing
StatePublished - 1987

ASJC Scopus subject areas

  • General Engineering

Fingerprint

Dive into the research topics of 'PROBLEM OF INTERNAL REPRESENTATIONS IN NEURAL NETS: CONCEPTS, IMPLICATIONS, AND A PROPOSED METRIC.'. Together they form a unique fingerprint.

Cite this