$\alpha$-Geodesical Skew Divergence
Information Theory
2021-04-27 v4 math.IT
Statistics Theory
Computation
Machine Learning
Statistics Theory
Abstract
The asymmetric skew divergence smooths one of the distributions by mixing it, to a degree determined by the parameter , with the other distribution. Such divergence is an approximation of the KL divergence that does not require the target distribution to be absolutely continuous with respect to the source distribution. In this paper, an information geometric generalization of the skew divergence called the -geodesical skew divergence is proposed, and its properties are studied.
Keywords
Cite
@article{arxiv.2103.17060,
title = {$\alpha$-Geodesical Skew Divergence},
author = {Masanari Kimura and Hideitsu Hino},
journal= {arXiv preprint arXiv:2103.17060},
year = {2021}
}