Title :
A Robust Extraction Algorithm Based on a Specific Kurtosis Value Range
Author :
Ye, Yalan ; Zhang, Zhi-Lin ; Chen, Jia ; Wu, Duojiao
Author_Institution :
Univ. of Electron. Sci. & Technol. of China, Chengdu
Abstract :
Independent component analysis (ICA), blind source separation (BSS) and related methods like blind source extraction (BSE) have been considered as a fundamental data analysis tool in the fields of neural network and signal processing. In this paper, we propose a robust algorithm based on a specific kurtosis value range that can extract a desired source signal as the first output signal with a specific kurtosis value range. That is to say, if we know that the kurtosis value of the desired signal generally lies in a specific range, while the values of other unwanted source signals do not belong to this range, we can use the proposed algorithm to extract the desired signal successfully. Moreover, the algorithm can work well in some poor situation (when the kurtosis values of some source signals are very close to each other). In addition, because of adopting an interior point penalty function method, the algorithm is robust to the estimation error of the kurtosis value range. Finally, we use the proposed algorithm to extract the accurate and reliable fetal electrocardiogram (FECG).
Keywords :
blind source separation; electrocardiography; feature extraction; independent component analysis; medical signal processing; BSE; BSS; ICA; blind source extraction; blind source separation; fetal electrocardiogram; independent component analysis; interior point penalty function method; kurtosis value range; robust extraction algorithm; unwanted source signals; Blind source separation; Data mining; Estimation error; Independent component analysis; Neural networks; Robustness; Signal processing; Signal processing algorithms; Source separation; Vectors;
Conference_Titel :
Natural Computation, 2007. ICNC 2007. Third International Conference on
Conference_Location :
Haikou
Print_ISBN :
978-0-7695-2875-5
DOI :
10.1109/ICNC.2007.131