Codivergences and information matrices
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
We propose a new concept of codivergence, which quantifies the similarity between two probability measures P1,P2 relative to a reference probability measure P0. In the neighborhood of the reference measure P0, a codivergence behaves like an inner product between the measures P1-P0 and P2-P0. Codivergences of covariance-type and correlation-type are introduced and studied with a focus on two specific correlation-type codivergences, the χ2-codivergence and the Hellinger codivergence. We derive explicit expressions for several common parametric families of probability distributions. For a codivergence, we introduce moreover the divergence matrix as an analogue of the Gram matrix. It is shown that the χ2-divergence matrix satisfies a data-processing inequality.
help_outline