摘要
作为ＥＢＬ（ＥｘｐｌａｎａｔｉｏｎＢａｓｅｄＬｅａｒｎｉｎｇ）的发展，从概念的自然形成过程出发，提出了一种新的概念模型ＦＥＢＭ（ＦｕｚｚｙＥｘｐｌａｎａｔｉｏｎＢａｓｅｄＭｏｄｅｌ）：当概念的解释谓词集为模糊集以及解释谓词取模糊逻辑值时，给出求概念真值的表达式；为了解决模糊概念的识别问题，引入了概念的模糊解释树ＦＥＴ．接着给出了对象的模糊识别算法ＦＥＢＬ．最后讨论了ＦＥＢＭ与ＦＥＢＬ的可操作性．
One of the most important problems in AI is the expression of knowledge concepts. Explanation based learning(EBL) is an important topic in AI, but it is not always practical to explain a knowledge concept very strictly. As development of EBL, according to natural process of concepts formation this paper presents a new concept model——fuzzy explanation based model (FEBM). When the explanation predicate set of concept is fuzzy and the truth value of the predicates themselves is fuzzy either, formal expressions is given to find truth value of “an object belonging to a concept”. In order to identify fuzzy objects, fuzzy explanation tree of concepts,FET, is presented. Then the paper presents an algorithm, FEBL, to identify object with fuzzy explanation. Finally the operationality of FEBM and FEBL is discussed.
出处
《计算机学报》
EI
CSCD
北大核心
1999年第6期615-619,共5页
Chinese Journal of Computers
关键词
知识概念
模糊模型
模糊目标识别
机器学习
Fuzzy explanation set, fuzzy number, move by weight, concept, fuzzy explanation tree(FET), fuzzy explanation based model of concept(FEBM).
