CuBICA: Independent component analysis by simultaneous third- and fourth-order cumulant diagonalization

被引：54

作者：

Blaschke, T ^{[1
]}

Wiskott, L ^{[1
]}

机构：

[1] Humboldt Univ, Inst Theoret Biol, Berlin, Germany

来源：

IEEE TRANSACTIONS ON SIGNAL PROCESSING | 2004年 / 52卷 / 05期

关键词：

contrast; cumulant; independent component analysis;

D O I：

10.1109/TSP.2004.826173

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

CuBICA, which is an improved method for independent component analysis (ICA) based on the diagonalization of cumulant tensors is proposed. It is based on Comon's algorithm, but it takes third- and fourth-order cumulant tensors into account simultaneously. The underlying contrast function is also mathematically much simpler and has a more intuitive interpretation. It is therefore easier to optimize and approximate. A comparison with Comon's and three other ICA algorithms on different data sets demonstrates its performance.

引用

页码：1250 / 1256

页数：7

共 19 条

[1] AMARI S, 1995, P INT S NONL THEOR I, P37
[2] [Anonymous], 2001, INDEPENDENT COMPONEN
[3] [Anonymous], 2002, Adaptive Blind Signal and Image Processing: Learning Algorithms and Applications
[4] AZZALINI A, 1985, SCAND J STAT, V12, P171
[5] AN INFORMATION MAXIMIZATION APPROACH TO BLIND SEPARATION AND BLIND DECONVOLUTION
BELL, AJ
SEJNOWSKI, TJ
[J]. NEURAL COMPUTATION, 1995, 7 (06) : 1129 - 1159
[6] BLIND BEAMFORMING FOR NON-GAUSSIAN SIGNALS
CARDOSO, JF
SOULOUMIAC, A
[J]. IEE PROCEEDINGS-F RADAR AND SIGNAL PROCESSING, 1993, 140 (06) : 362 - 370
[7] High-order contrasts for independent component analysis
Cardoso, JF
[J]. NEURAL COMPUTATION, 1999, 11 (01) : 157 - 192
[8] INDEPENDENT COMPONENT ANALYSIS, A NEW CONCEPT
COMON, P
[J]. SIGNAL PROCESSING, 1994, 36 (03) : 287 - 314
[9] COMON P, 1994, P 10 IFAC S SYST ID, V1, P77
[10] COMON P, 2002, P 2 IMA C MATH COMM

← 1 2 →