Local Independent Component Analysis with Fuzzy Clustering and Regression-principal Component Analysis

Independent component analysis (ICA) is an unsupervised technique for blind source separation, and the ICA algorithms using non-gaussianity as the measure of mutual independence have been also used for projection pursuit or visualization for knowledge discovery in databases (KDD). However, in real applications, it is often the case that we fail to extract useful latent variables because they have no connection with predefined criterion variables. This paper proposes an enhanced technique of ICA, which extracts independent components closely related to some external criteria. Preprocessing is performed by using fuzzy regression-principal component analysis, which estimates latent variables that have high correlation with the external criteria considering local data structure.