Knowledge Base Construction of Ship Design Using Fuzzy Equalization and Rough Sets |
Suh Kyu-Youl |
Department of Naval Architecture Dong-Myung Univ. |
퍼지균등화와 러프집합을 이용한 선박설계 지식기반 구축 |
서규열 |
동명대학교 조선공학과 |
|
© 2007 The Korean Society of Ocean Engineers
Open access / Under a Creative Commons License
|
Keywords:
Automatic rule derivation, Fuzzy equalization, Rough sets, Consistent measure, Expert system |
핵심용어:
자동규칙추출, 퍼지균등화, 러프집합, 일관성척도, 전문가시스템 |
Abstract |
Inference rules of the knowledge base, generated by experts or optimization, may be often inconsistent and incomplete. This paper suggests a systematic and automatic method which extracts inference rules not from experts' subject but from data. First, input/output linguistic variables are partitioned into several properties by the fuzzy equalization algorithm and each combination of their properties comes to premise of inference rule. Then, the conclusion which is the mast suitable for the premise is selected by evaluating consistent measure. This method, automatically from data, derives inference rules from experience. It is shown through application that extracts new inference rules between hull dimensions and hull performance. |
|