一种基于神经网络的规则获取方法

Based on Neural Network for Rule Extraction Approach

分类问题是数据挖掘中的一个重要问题 ,分类目的就是寻找规则 ,具体来说 ,就是从给定的数据集合中找出能把数据集划分成不相交的若干个组的规则 ,目前已有的在大型数据库中挖掘分类规则的数据挖掘方法 ,主要还是基于符号学习机制的决策树方法 .本文研究了一种新型的规则抽取算法 。

Classification is one class of data mining problems, which involves finding rules that partition a given data set into disjoint groups. Approach proposed so far for mining classification rules for large databases are mainly decision tree based symbolic learning methods. In this paper, we study a new method to extract rules from neural networks.