The Use of Rough Sets Theory for Clustering Rule Extraction of ART 1
In order to find the rules that can deduce the clustering results of ART1, we try to extract them by using rough sets. In this paper, firstly, we describe the input-output mapping relationship of ART1 by information table in rough sets theory. Secondly, we reduce the attributes, extract rules and calculate the confidence degrees of rules extracted. The decision rules obtained can reconstruct the input-output mapping relationship of ART1. Finally, we combined sets of clustering rules corresponding to different ART1s, and get a set of new rules, which under certain condition can deduce the clustering result for other ART1s.
Neural networks ART1 Rough sets Clustering rules
Xiaoqin Zhu Shenchu Xu Zhenxiang Chen Haixin Chen
Department of physics, Xiamen University, Xiamen 361005, China
国际会议
8th International Conference on Neural Information Processing(ICONIP 2001)(第八届国际神经信息处理大会)
上海
英文
393-398
2001-11-14（万方平台首次上网日期，不代表论文的发表时间）