A Filter Approach to Feature Selection Based on Mutual Information

摘要：

In pattern recognition, feature selection aims to choose the smallest subset of features that is necessary and sufficient to describe the target concept. In this paper, a mutual information-based constructive criterion under arbitrary information distributions of input features is presented for feature selection. This criterion can capture both the relevance to the output classes and the redundancy with respect to the already-selected features without any parameters like ft in MIPS or MIFS-U methods to be preset Furthermore, a modified greedy feature selection algorithm called MICC is proposed, and experimental results demonstrate the good performance of MICC on both synthetic and benchmark data sets.

关键词： Pattern classification machine learning feature selection filter approach mutual information.

作者: Jinjie Huang Yunze Cai Xiaoming Xu

作者单位: Department of Automation, Shanghai Jiao Tong University, Dongchuan Road 800, Shanghai 200240, China

会议类型: 国际会议

会议名称: Firth IEEE International Conference on Cognitive Informatics(第五届认知信息国际会议)

会议地点: 北京

会议语种:英文

页码: 84-89

在线出版日期: 2006-07-17（万方平台首次上网日期，不代表论文的发表时间）