arXiv Analytics

Sign in

arXiv:2005.02196 [cs.LG]AbstractReferencesReviewsResources

Measuring the Discrepancy between Conditional Distributions: Methods, Properties and Applications

Shujian Yu, Ammar Shaker, Francesco Alesiani, Jose C. Principe

Published 2020-05-05Version 1

We propose a simple yet powerful test statistic to quantify the discrepancy between two conditional distributions. The new statistic avoids the explicit estimation of the underlying distributions in highdimensional space and it operates on the cone of symmetric positive semidefinite (SPS) matrix using the Bregman matrix divergence. Moreover, it inherits the merits of the correntropy function to explicitly incorporate high-order statistics in the data. We present the properties of our new statistic and illustrate its connections to prior art. We finally show the applications of our new statistic on three different machine learning problems, namely the multi-task learning over graphs, the concept drift detection, and the information-theoretic feature selection, to demonstrate its utility and advantage. Code of our statistic is available at https://bit.ly/BregmanCorrentropy.

Comments: accepted at IJCAI 20, code is available at https://github.com/SJYuCNEL/Bregman-Correntropy-Conditional-Divergence
Categories: cs.LG, cs.IT, math.IT, stat.ML
Related articles: Most relevant | Search more
arXiv:1201.5338 [cs.LG] (Published 2012-01-25, updated 2012-09-21)
On Constrained Spectral Clustering and Its Applications
arXiv:1812.08434 [cs.LG] (Published 2018-12-20)
Graph Neural Networks: A Review of Methods and Applications
arXiv:1905.00331 [cs.LG] (Published 2019-05-01)
High-Performance Support Vector Machines and Its Applications