计算机科学 ›› 2016, Vol. 43 ›› Issue (6): 122-126.doi: 10.11896/j.issn.1002-137X.2016.06.025
顾玲玲,刘国庆
GU Ling-ling and LIU Guo-qing
摘要: 独立成分分析(Independent Component Analysis,ICA)是解决盲源分离问题十分有效的方法。特别是Fast-ICA算法,它以中心极限定理为出发点,采用定点迭代的优化算法,收敛快速、稳健。但是在提取弱信号时,由于中心极限定理不再严格成立,FastICA算法也不再适用。因此从理论和实验两个方面着手验证了这个观点,并针对弱信号提取问题提出新的解决思路:在FastICA算法的基础上,引入源信号的部分先验信息作为约束,即参考独立成分分析(Independent Component Analysis with Reference,ICA-R)。若已知源信号的部分功率谱,结合加权范数最小化信号外推算法的思想,建立接近性度量,以约束的形式融入FastICA算法中,从而分离出要求的弱信号。实验结果表明,不管是对模拟信号还是真实的脑电信号,该算法都是有效的。
[1] Hyvrinen A.Fast and robust fixed-point algorithms for independent component analysis[J].IEEE Transactions on Neural Networks,1999,0(3):626-634 [2] Cardoso J F,Souloumiac A.Blind beamforming for non-Gaussiansignals[J].IEEE Proceedings F (Radar and Signal Processing),1993,140(6):362-370 [3] Wang J,Zhou J,Peng B.Weak signal detection method based on Duffing oscillator[J].Kybernetes,2009,38(10):1662-1668 [4] Zhang Z L,Yi Z.Extraction of a source signal whose kurtosis value lies in a specific range[J].Neurocomputing,2006,69(7):900-904 [5] Zhang Z L.Morphologically constrained ICA for extracting weaktemporally correlated signals[J].Neurocomputing,2008,71(7):1669-1679 [6] 杨福生,洪波.独立分量分析的原理与应用[M].北京:清华大学出版社,2005 [7] Hyvrinen A,Erkki O.Independent component analysis:Algorithms and Applications[J].Neural Networks,2000,13(4/5):411-430 [8] 李贤平.概率论基础[M].北京:高等教育出版社,1997,10(3):626-634 [9] Lu W,Rajapakse J C.ICA with reference[J].Neurocomputing,2006,69(16-18):244-256 [10] Lu W,Rajapakse J C.ICA with reference[C]∥Proc.Third Int.Conf on ICA and Blind source separation (ICA2001).2001:120-125 [11] Cabrera S D,Parks T W.Extrapolation and spectral estimation with iterative weighted norm modification[J].IEEE Transactions on Signal Processing,1991,9(4):842-850 [12] 解可新,韩健,林友联.最优化方法(修订版)[M].天津:天津大学出版社,2004 [13] Bear M F,Connors B W,Paradiso M A.Neuroscience:Exploring the Brain(第二版)[M].王建军,主译.北京:高等教育出版社,2004:578-582 [14] Ito J,Roy S,Liu Y,et al.Whisker barrel cortex delta oscillations and gamma power in the awake mouse are linked to respiration[D].Nature Communication,2014 [15] 倪国熙.常用的矩阵理论和方法[M].上海:上海科学技术出版社,1984 |
No related articles found! |
|