On extending the complex FastICA algorithms to noisy data

Zongli Ruan; Liping Li; Guobing Qian

doi:10.1016/j.neunet.2014.08.013

On extending the complex FastICA algorithms to noisy data

Neural Netw. 2014 Dec:60:194-202. doi: 10.1016/j.neunet.2014.08.013. Epub 2014 Sep 8.

Authors

Zongli Ruan¹, Liping Li², Guobing Qian²

Affiliations

¹ School of Electronic Engineering, University of Electronics Science and Technology of China, Chengdu 611731, China; College of Science, China University of Petroleum, Qingdao 266580, China. Electronic address: ruanzl0496@gmail.com.
² School of Electronic Engineering, University of Electronics Science and Technology of China, Chengdu 611731, China.

PMID: 25243358
DOI: 10.1016/j.neunet.2014.08.013

Abstract

Independent component analysis (ICA) methods are widely applied to modern digital signal processing. The complex-valued FastICA algorithms are one type of the most significant methods. However, the complex ICA model usually omits the noise. In this paper, we discuss two complex FastICA algorithms for noisy data, where the cost functions are based on kurtosis and negentropy respectively. The nc-FastICA and KM-F algorithms are modified to separate noisy data. At the same time, we also give the stability conditions of cost functions. Simulations are presented to illustrate the effectiveness of our methods.

Keywords: Complex fast fixed-point algorithm; Independent component analysis (ICA); Noisy data; Pseudo-whitening; Stability condition.

MeSH terms

Algorithms*
Signal Processing, Computer-Assisted*
Signal-To-Noise Ratio*