Towards full automation of the discovery of heuristics in a nuclear engineering project: Integration with a neural information language

Ephraim Nissan, Hava Siegelmann, Alex Galperin, Shuky Kimhi

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

7 Scopus citations

Abstract

FUELCON is an expert system in nuclear engineering. The practitioner of nuclear fuel management uses it to generate configurations of reactor core refueling. The domain expert, on the other hand, employs the system to discover new heuristics for the previous task, based on performance during previous iterations in the same session. Expert use involves a manual phase of revising the ruleset. We expose the building blocks of the design of a new version, that incorporates a neural component to carry out the revision. It uses the system’s previous performance it for adaptation and learning better rules. for adaptation and learning better rules. The neural component is based on a particular language and schema for symbolic to recurrent-analog conversion, called NEL, and on reinforcement learning for the adaptation.

Original languageEnglish
Title of host publicationMethodologies for Intelligent Systems - 8th International Symposium, ISMIS 1994, Proceedings
EditorsZbigniew W. Ras, Maria Zemankova
PublisherSpringer Verlag
Pages428-436
Number of pages9
ISBN (Print)9783540584957
DOIs
StatePublished - 1 Jan 1994
Event8th International Symposium on Methodologies for Intelligent Systems, ISMIS 1994 - Charlotte, United States
Duration: 16 Oct 199419 Oct 1994

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume869 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference8th International Symposium on Methodologies for Intelligent Systems, ISMIS 1994
Country/TerritoryUnited States
CityCharlotte
Period16/10/9419/10/94

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'Towards full automation of the discovery of heuristics in a nuclear engineering project: Integration with a neural information language'. Together they form a unique fingerprint.

Cite this