N°24-50: Adaptive Joint Distribution Learning

AuthorD. Filipović, P. Schneider, M. Multerer
Date24 sept. 2024
CatégorieWorking Papers

We develop a new framework for estimating joint probability distributions using tensor product reproducing kernel Hilbert spaces (RKHS). Our framework accommodates a low-dimensional, normalized and positive model of a Radon-Nikodym derivative, which we estimate from sample sizes of up to several millions, alleviating the inherent limitations of RKHS modeling. Well-defined normalized and positive conditional distributions are natural by-products to our approach. Our proposal is fast to compute and accommodates learning problems ranging from prediction to classification. Our theoretical findings are supplemented by favorable numerical results.

Notre site utilise des cookies afin de vous garantir la meilleure expérience possible. En navigant sur notre site, vous acceptez notre utilisation de cookies.
RefuserAccepter