Generalizing Independent Component Analysis for Two Related Data Sets
Juha Karhunen, Tomas Ukkonen · The 2006 IEEE International Joint Conference on Neural Network Proceedings · 2006
We introduce in this paper methods for finding mutually corresponding dependent components from two different but related data sets in an unsupervised (blind) manner. The basic idea is to generalize cross-correlation analysis for taking into account higher-order statistics. We propose independent component analysis (ICA) type extensions for the singular value decomposition of the cross-correlation matrix. They extend cross-correlation analysis in a similar manner as ICA extends standard principal component analysis for covariance matrices. We present experimental results demonstrating the usefulness of the proposed methods both for artificially generated data and for a cryptographic problem.