Cauchy Schwarz Divergence

Cauchy-Schwarz (CS) divergence is a statistical measure quantifying the difference between probability distributions, offering advantages over traditional metrics like Kullback-Leibler divergence, particularly in scenarios involving multiple distributions or conditional probabilities. Current research focuses on extending CS divergence to handle multiple distributions (generalized CS divergence) and conditional distributions (conditional CS divergence), finding applications in diverse machine learning tasks such as domain adaptation, clustering, and sequential decision-making. These advancements provide robust and computationally efficient tools for comparing distributions, improving the performance of various algorithms and broadening the applicability of deep learning models across different domains.

Papers

May 30, 2024

Domain Adaptation with Cauchy-Schwarz Divergence
Wenzhe Yin, Shujian Yu, Yicong Lin, Jie Liu, Jan-Jakob Sonke, Efstratios Gavves
Domain Adaptation Unsupervised Domain Adaptation Inverse Divergence Generalization Error Bound Cauchy Schwarz Divergence

May 7, 2024

Generalized Cauchy-Schwarz Divergence and Its Deep Learning Applications
Mingfei Lu, Chenxu Li, Shujian Yu, Robert Jenssen, Badong Chen
Deep Learning Application Inverse Divergence Multiple Distribution Divergence Measure Cauchy Schwarz Divergence

January 21, 2023

The Conditional Cauchy-Schwarz Divergence with Applications to Time-Series Data and Sequential Decision Making
Shujian Yu, Hongming Li, Sigurd Løkse, Robert Jenssen, José C. Príncipe
Financial Application Time Series Data Conditional Reasoning Sequential Decision Making Maximum Mean Discrepancy Inverse Divergence KL Divergence Cauchy Schwarz Divergence

Cauchy Schwarz Divergence

Papers

Domain Adaptation with Cauchy-Schwarz Divergence

Generalized Cauchy-Schwarz Divergence and Its Deep Learning Applications

The Conditional Cauchy-Schwarz Divergence with Applications to Time-Series Data and Sequential Decision Making