Complex infomax: Convergence and approximation of Infomax with complex nonlinearities

被引:7
作者
Calhoun, Vince
Adali, Tuelay
机构
[1] Inst Living, Olin Neuropsychiat Res Ctr, Hartford, CT 06106 USA
[2] Yale Univ, Dept Psychiat, New Haven, CT 06520 USA
[3] Univ Maryland Baltimore Cty, Dept CSEE, Baltimore, MD 21250 USA
来源
JOURNAL OF VLSI SIGNAL PROCESSING SYSTEMS FOR SIGNAL IMAGE AND VIDEO TECHNOLOGY | 2006年 / 44卷 / 1-2期
基金
美国国家科学基金会;
关键词
fMRI; Infomax; ICA; independent component analysis;
D O I
10.1007/s11265-006-7514-5
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Independent component analysis (ICA) for separating complex-valued sources is needed for convolutive source-separation in the frequency domain, or for performing source separation on complex-valued data, such as functional magnetic resonance imaging or radar data. Previous complex Infomax approaches that use nonlinear functions in the updates have proposed using bounded (and hence non-analytic) nonlinearities. In this paper, we propose using an analytic (and hence unbounded) complex nonlinearity for Infomax for processing complex-valued sources. We show by simulation examples that using an analytic nonlinearity for processing complex data has a number of advantages. First, when compared to split-complex approaches (i.e., approaches that split the real and imaginary data into separate channels), the shape of the performance surface is improved resulting in better convergence characteristics. We also show that using an analytic complex-valued function for the nonlinearity is more effective in generating the higher order statistics required to establish independence when compared to complex nonlinear functions, i.e., functions that are C -> C.
引用
收藏
页码:173 / 190
页数:18
相关论文
共 43 条
[1]  
ADALI T, 2004, P ICASSP MONTR CAN
[2]  
Amari S, 1996, ADV NEUR IN, V8, P757
[3]   Complex independent component analysis of frequency-domain electroencephalographic data [J].
Anemüller, J ;
Sejnowski, TJ ;
Makeig, S .
NEURAL NETWORKS, 2003, 16 (09) :1311-1323
[4]  
ANNEMUELLER J, 2003, P INT C ICA BSS
[5]  
BACK A, 1994, P NNSP ED SCOTL
[6]   AN INFORMATION MAXIMIZATION APPROACH TO BLIND SEPARATION AND BLIND DECONVOLUTION [J].
BELL, AJ ;
SEJNOWSKI, TJ .
NEURAL COMPUTATION, 1995, 7 (06) :1129-1159
[7]   ON THE COMPLEX BACKPROPAGATION ALGORITHM [J].
BENVENUTO, N ;
PIAZZA, F .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 1992, 40 (04) :967-968
[8]  
Bingham E, 2000, Int J Neural Syst, V10, P1, DOI 10.1142/S0129065700000028
[9]   Independent component analysis of fMRI data in the complex domain [J].
Calhoun, VD ;
Adali, T ;
Pearlson, GD ;
van Zijl, PCM ;
Pekar, JJ .
MAGNETIC RESONANCE IN MEDICINE, 2002, 48 (01) :180-192
[10]   Spatial and temporal independent component analysis of functional MRI data containing a pair of task-related waveforms [J].
Calhoun, VD ;
Adali, T ;
Pearlson, GD ;
Pekar, JJ .
HUMAN BRAIN MAPPING, 2001, 13 (01) :43-53