A NOVEL ENTROPY ESTIMATOR AND ITS APPLICATION TO leA
Tiilay Adali · 2009
We present a new (differential) entropy estimator where the maximum entropy bound is used to approximate the entropy given the observations, and is computed using a numerical procedure. The resulting accurate estimate for the entropy is used to derive a new algorithm to perform independent com ponent analysis (ICA). The new algorithm, ICA by entropy bound minimization (ICA-EBM), adopts a line search pro cedure, and initially uses updates that constrain the demix ing matrix to be orthogonal for robust performance. We present simulation results that demonstrate the superior per formance of ICA-EBM and its ability to match sources that come from a wide range of distributions.