Independent Component Analysis Using Convex Divergence

摘要：

The convex divergence is used as a surrogate function for obtaining a class of 1CA algorithms (Independent Component Analysis) called the f-ICA. The convex divergence is a super class of a-divergence, which is a further upper family of Kullback-Leibler divergence or mutual information. Therefore, the f-ICA contains the a-lCA and the minimum mutual information ICA. In addition to theoretical interest of generalization, the f-ICA contains a subset faster than the minimum mutual information ICA. It is found that this speed control is equivalent to the α-ICA. Finally, applications to brain fMRI maps distillation is presented.

作者: Yasuo Matsuyama Naoto Katsumata Shuichiro Imahara

作者单位: Department of Electrical, Electronics and Computer Engineering, Waseda University, Tokyo 169-8555, Japan

会议类型: 国际会议

会议名称: 8th International Conference on Neural Information Processing(ICONIP 2001)(第八届国际神经信息处理大会)

会议地点: 上海

会议语种:英文

页码: 1211-1216

在线出版日期: 2001-11-14（万方平台首次上网日期，不代表论文的发表时间）