NONLINEAR INDEPENDENT COMPONENT ANALYSIS USING ENSEMBLE LEARNING: EXPERIMENTS AND DISCUSSION

Harri Valpola 1 , Xavier Giannakopoulos 2 , Antti Honkela 1 , and Juha Karhunen 1, Juha.Karhunen@hut.fi URL: http://www.cis.hut.fi/
E≠mail: Harri.Valpola@hut.fi, Xavier@idsia.ch, Antti.Honkela@hut.fi,, Juha.Karhunen@hut.fi URL: http://www.cis.hut.fi/

In this paper, we present experimental results on a non≠ linear independent component analysis approach based on Bayesian ensemble learning. The theory of the method is discussed in a companion paper. Simulations with artiÝcial and natural data demonstrate the feasi≠ bility and good performance of the proposed approach. We also discuss the relationships of the method to other existing methods.