|
摘要：
|
|
基于事例的推理技术 CBR是人工智能技术发展的重要领域之一 ,而大型事例库的检索 ,一般的索引法和最近邻法等方法难以达到快速、准确的检索要求。因此 ,有必要研究智能检索推理技术。本文根据事例学习的决策树 ID3和 PID算法 ,提出了多概念学习决策树 MNID算法 ,开发了基于事例学习的智能模具相似结构推理系统 ,并应用于冲模凸模结构 Case库的检索 ,效果很好 ,系统通过事例的自主学习 ,产生出推理规则 ,并能根据用户对所设计模具结构的要求 ,自动检索出符合设计要求的相似 Case,实现了 CAD过程中大型 Case库的自动检索。
|
|
CBR (case based reasoning) is one of the important techniques on development of artificial intelligent field.It is difficult for designer to obtain the similar cases quickly by general retrieval method in big case\|base.Therefore,It is necessary to research artificial intelligent retrieving technique.According to ID3 and PID decision tree algorithm based on cases learning,the new decision tree algorithm MNID had been advanced in this paper,which was an algorithm to consider multiple concept learning,and the reasoning system of die & mold structure case based on case learning was developed.It was effective to apply the system to die similar punch structure case retrieval.The reasoning rules could be obtained by automatic cases learning in this system.The similar cases of meeting designer's demand could be retrieved out.And the automatic retrieval for big case\|base was realized by the means of MNID algorithm.
|
|
基金项目：
|
|
中国博士后基金
|
|
作者简介：
|
|
|
|
参考文献：
|
1 JR Quinlan.Induction of decision tree.Machine Learning,1 986( 1 ) :81～ 1 0 6
2 洪家荣等 .一种新的决策树归纳学习算法 .计算机学报 ,1 995( 6) :4 70～ 4 74
|
|
服务与反馈：
|
|
【文章下载】【加入收藏】
|
|
|
|
