Enregistré dans:
Détails bibliographiques
Auteurs principaux: Cha, Jinho, Kim, Youngchul, Shin, Jungmin, Cho, Jaeyoung, Kim, Seon Jin, Ryu, Junyeol
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2510.20227
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866915571560349696
author Cha, Jinho
Kim, Youngchul
Shin, Jungmin
Cho, Jaeyoung
Kim, Seon Jin
Ryu, Junyeol
author_facet Cha, Jinho
Kim, Youngchul
Shin, Jungmin
Cho, Jaeyoung
Kim, Seon Jin
Ryu, Junyeol
contents We develop a general optimization-theoretic framework for Bregman-Variational Learning Dynamics (BVLD), a new class of operator-based updates that unify Bayesian inference, mirror descent, and proximal learning under time-varying environments. Each update is formulated as a variational optimization problem combining a smooth convex loss f_t with a Bregman divergence D_psi. We prove that the induced operator is averaged, contractive, and exponentially stable in the Bregman geometry. Further, we establish Fejer monotonicity, drift-aware convergence, and continuous-time equivalence via an evolution variational inequality (EVI). Together, these results provide a rigorous analytical foundation for well-posed and stability-guaranteed operator dynamics in nonstationary optimization.
format Preprint
id arxiv_https___arxiv_org_abs_2510_20227
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Optimization of Bregman Variational Learning Dynamics
Cha, Jinho
Kim, Youngchul
Shin, Jungmin
Cho, Jaeyoung
Kim, Seon Jin
Ryu, Junyeol
Optimization and Control
We develop a general optimization-theoretic framework for Bregman-Variational Learning Dynamics (BVLD), a new class of operator-based updates that unify Bayesian inference, mirror descent, and proximal learning under time-varying environments. Each update is formulated as a variational optimization problem combining a smooth convex loss f_t with a Bregman divergence D_psi. We prove that the induced operator is averaged, contractive, and exponentially stable in the Bregman geometry. Further, we establish Fejer monotonicity, drift-aware convergence, and continuous-time equivalence via an evolution variational inequality (EVI). Together, these results provide a rigorous analytical foundation for well-posed and stability-guaranteed operator dynamics in nonstationary optimization.
title Optimization of Bregman Variational Learning Dynamics
topic Optimization and Control
url https://arxiv.org/abs/2510.20227