摘要
为了提高实例推理系统中实例检索的效率与质量,提出了基于神经网络的、和最近邻相似度相结合的实例检索算法。首先,利用概率神经网络实现对实例的动态分类,以缩小实例检索范围;其次,介绍了实例属性相似度计算方法,针对各属性相似度的权重问题,给出主观和客观赋权方法;最后,给出截止阀总体设计方案生成的算例,验证方法的可行性。
In order to improve the efficiency and quality of case retrieval in case-based reasoning system,a case retrieval algorithm based on the combination of the neural network and the nearest neighbor similarity is present.Firstly,dynamic classification of cases was conducted through probabilistic neural network so as to shrink the casesearching range.Secondly,the property similarity algorithm is introduced.Then for the weight assignments,the subjective and objective weight assignment methods are introduced.Finally,an example of the cut-off valve design scheme is given to illustrate validity.
出处
《机电一体化》
2014年第11期63-67,共5页
Mechatronics
关键词
实例检索
神经网络
相似度
权重
case retrieval
neural network
similarity
weight
