Cheeger Cut Model for the Balanced Data Classification Problem

摘要：

　　In this paper we propose a numerical method based on the splitting strategy to solve the Cheeger cut model.In order to improve the classification results,we propose a new self-tuning strategy to choose a robust scaling parameter.Some numerical examples are arranged to illustrate the efficiency of our proposed method.

关键词： Cheeger Cut Penalty Method Weighted Function Scaling Parameter

作者: Yan-zhou Zhang Yan Jiang Zhi-Feng Pang

作者单位: Basic Courses Department, Henan Polytechnic, Zhengzhou, 450046, China Foundation department, Zhengzhou Tourism College, Zhengzhou, 45009, China College of Mathematics and Information Science, Henan University, Kaifeng, 475004, China

会议类型: 国际会议

会议名称: 2013 2nd International Conference on Systems Engineering and Modeling(ICSEM-13)(2013年第二届系统工程与建模国际会议)

会议地点: 北京

会议语种:英文

页码: 694-698

在线出版日期: 2013-04-19（万方平台首次上网日期，不代表论文的发表时间）