An algorithm for real-time independent component analysis in dynamic environments
Shuxue Ding, Daming Wei, Sadao Omata · 2004
We present a novel algorithm for independent component analysis (ICA) based on gradient learning with simultaneous perturbation stochastic approximation (SPSA). This algorithm can work well in on-line mode of ICA processing, in a dynamic mixing environment. It converges very fast even for non-stationary, and/or non-identically independent distributed (non-I.I.D.) signals, so that the algorithm is very suitable for most real-time applications. In this paper, theories and implementations of the algorithm are described. Results of computer simulation are also presented to demonstrate the effectiveness.