AUTOMATIC RULE EXTRACTION FROM STATISTICAL DATA AND FUZZY TREE SEARCH.

Shigeo Morishima, Hiroshi Harashima

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

This paper proposes an automatic rule extraction mechanism using statistical analysis. In this system, production rules are expressed in the form of a threshold function, which is expressed easily as a linear combination of input vectors and weighting coefficients. Thus weighting coefficients can be calculated by the same method as a discriminant function. If only one threshold is defined, general Boolean logic can be expressed. An ambiguous inference rule can be expressed when the threshold levels are multidefined and a membership function is defined at each category. A Fuzzy Tree Search algorithm which combines ambiguous interference and tree search is proposed. This algorithm can search and determine an optimum cluster with little calculation and good performance. A medical diagnostic system for psychiatry has been constructed based on this algorithm and inference rules have been extracted automatically. This experiment shows that Fuzzy Tree Search is as fast as the tree search technique and has as good a performance as a full search clustering technique.

Original languageEnglish
Pages (from-to)26-37
Number of pages12
JournalSystems and Computers in Japan
Volume19
Issue number5
Publication statusPublished - 1988 May
Externally publishedYes

Fingerprint

Rule Extraction
Search Trees
Inference Rules
Ambiguous
Weighting
Production Rules
Discriminant Function
Threshold Function
Trees (mathematics)
Tree Algorithms
Coefficient
Membership functions
Membership Function
Search Algorithm
Statistical Analysis
Linear Combination
Statistical methods
Diagnostics
Interference
Clustering

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Hardware and Architecture
  • Information Systems
  • Theoretical Computer Science

Cite this

AUTOMATIC RULE EXTRACTION FROM STATISTICAL DATA AND FUZZY TREE SEARCH. / Morishima, Shigeo; Harashima, Hiroshi.

In: Systems and Computers in Japan, Vol. 19, No. 5, 05.1988, p. 26-37.

Research output: Contribution to journalArticle

@article{40c6909900d5447da57c07531f106d8b,
title = "AUTOMATIC RULE EXTRACTION FROM STATISTICAL DATA AND FUZZY TREE SEARCH.",
abstract = "This paper proposes an automatic rule extraction mechanism using statistical analysis. In this system, production rules are expressed in the form of a threshold function, which is expressed easily as a linear combination of input vectors and weighting coefficients. Thus weighting coefficients can be calculated by the same method as a discriminant function. If only one threshold is defined, general Boolean logic can be expressed. An ambiguous inference rule can be expressed when the threshold levels are multidefined and a membership function is defined at each category. A Fuzzy Tree Search algorithm which combines ambiguous interference and tree search is proposed. This algorithm can search and determine an optimum cluster with little calculation and good performance. A medical diagnostic system for psychiatry has been constructed based on this algorithm and inference rules have been extracted automatically. This experiment shows that Fuzzy Tree Search is as fast as the tree search technique and has as good a performance as a full search clustering technique.",
author = "Shigeo Morishima and Hiroshi Harashima",
year = "1988",
month = "5",
language = "English",
volume = "19",
pages = "26--37",
journal = "Systems and Computers in Japan",
issn = "0882-1666",
publisher = "John Wiley and Sons Inc.",
number = "5",

}

TY - JOUR

T1 - AUTOMATIC RULE EXTRACTION FROM STATISTICAL DATA AND FUZZY TREE SEARCH.

AU - Morishima, Shigeo

AU - Harashima, Hiroshi

PY - 1988/5

Y1 - 1988/5

N2 - This paper proposes an automatic rule extraction mechanism using statistical analysis. In this system, production rules are expressed in the form of a threshold function, which is expressed easily as a linear combination of input vectors and weighting coefficients. Thus weighting coefficients can be calculated by the same method as a discriminant function. If only one threshold is defined, general Boolean logic can be expressed. An ambiguous inference rule can be expressed when the threshold levels are multidefined and a membership function is defined at each category. A Fuzzy Tree Search algorithm which combines ambiguous interference and tree search is proposed. This algorithm can search and determine an optimum cluster with little calculation and good performance. A medical diagnostic system for psychiatry has been constructed based on this algorithm and inference rules have been extracted automatically. This experiment shows that Fuzzy Tree Search is as fast as the tree search technique and has as good a performance as a full search clustering technique.

AB - This paper proposes an automatic rule extraction mechanism using statistical analysis. In this system, production rules are expressed in the form of a threshold function, which is expressed easily as a linear combination of input vectors and weighting coefficients. Thus weighting coefficients can be calculated by the same method as a discriminant function. If only one threshold is defined, general Boolean logic can be expressed. An ambiguous inference rule can be expressed when the threshold levels are multidefined and a membership function is defined at each category. A Fuzzy Tree Search algorithm which combines ambiguous interference and tree search is proposed. This algorithm can search and determine an optimum cluster with little calculation and good performance. A medical diagnostic system for psychiatry has been constructed based on this algorithm and inference rules have been extracted automatically. This experiment shows that Fuzzy Tree Search is as fast as the tree search technique and has as good a performance as a full search clustering technique.

UR - http://www.scopus.com/inward/record.url?scp=0024017123&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0024017123&partnerID=8YFLogxK

M3 - Article

VL - 19

SP - 26

EP - 37

JO - Systems and Computers in Japan

JF - Systems and Computers in Japan

SN - 0882-1666

IS - 5

ER -