Skip to main content
Journal of Machine Learning Research (2003)
  • Erik G Learned-Miller, University of Massachusetts - Amherst
  • John W Fisher, III, Massachusetts Institute of Technology

This paper presents a new algorithm for the independent components analysis (ICA) problem based on an efficient entropy estimator. Like many previous methods, this algorithm directly minimizes the measure of departure from independence according to the estimated Kullback-Leibler divergence between the joint distribution and the product of the marginal distributions. We pair this approach with efficient entropy estimators from the statistics literature. In particular, the entropy estimator we use is consistent and exhibits rapid convergence. The algorithm based on this estimator is simple, computationally efficient, intuitively appealing, and outperforms other well known algorithms. In addition, the estimator’s relative insensitivity to outliers translates into superior performance by our ICA algorithm on outlier tests. We present favorable comparisons to the Kernel ICA, FAST-ICA, JADE, and extended Infomax algorithms in extensive simulations. We also provide public domain source code for our algorithms.

Publication Date
December, 2003
Citation Information
Erik G Learned-Miller and John W Fisher. "ICA USING SPACINGS ESTIMATES OF ENTROPY" Journal of Machine Learning Research Vol. 4 (2003)
Available at: