Local Independent Component Analysis Using Clustering
Juha Karhunen, S. Malaroiu · 1999
In standard ICA, a linear data model is used for a global description of the data. Even though linear ICA yields meaningful results in many cases, it can provide a crude approximation only for nonlinear data distributions. We propose a new structure, where local ICA models are used in connection with a suitable clustering algorithm grouping the data. The clustering part is responsible for an overall coarse nonlinear representation of the underlying data, while linear ICA models of each cluster are used for describing local features of the data. The goal is to represent the data better than in linear ICA while avoiding computational difficulties related with nonlinear ICA. We discuss connections to existing methods, and present experimental results for natural image data.