A Robust Approach for ICA with High-level Noisy Measurements and Unknown Number of Sources

摘要：

This paper presents a robust approach for ICA based on the subspace method and parameterized t-distribution density model. This approach includes several techniques; pre-whitenmg with noise reduction, optimal dimensionality reduction (estimation of unknow source number), and the decomposition of the mixtures of sub-Gaussian and superGaussian source components. The results illustrate that the effectiveness of this robust approach.

作者: Jianting Cao

作者单位: Dept. of Electrical and Electronics Engineering, Sophia University,7-1 Kioicho, Chiyoda-ku, Tokyo 102-8554, Japan;Brain Science Institute, RIKEN 2-1 Hirosawa, Wako-shi, Saitama 351-0198, 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: 1050-1055

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）