Generalized Kernelized Bandits: A Novel Self-Normalized Bernstein-Like Dimension-Free Inequality and Regret Bounds

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Metelli, Alberto Maria, Drago, Simone, Mussi, Marco
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914193448370176
author Metelli, Alberto Maria
Drago, Simone
Mussi, Marco
author_facet Metelli, Alberto Maria
Drago, Simone
Mussi, Marco
contents We study the regret minimization problem in the novel setting of generalized kernelized bandits (GKBs), where we optimize an unknown function $f^*$ belonging to a reproducing kernel Hilbert space (RKHS) having access to samples generated by an exponential family (EF) reward model whose mean is a non-linear function $μ(f^*)$. This setting extends both kernelized bandits (KBs) and generalized linear bandits (GLBs), providing a unified view of both settings. We propose an optimistic regret minimization algorithm, GKB-UCB, and we explain why existing self-normalized concentration inequalities used for KBs and GLBs do not allow to provide tight regret guarantees. For this reason, we devise a novel self-normalized Bernstein-like dimension-free inequality that applies to a Hilbert space of functions with bounded norm, representing a contribution of independent interest. Based on it, we analyze GKB-UCB, deriving a regret bound of order $\widetilde{O}( γ_T \sqrt{T/κ_*})$, being $T$ the learning horizon, $γ_T$ the maximal information gain, and $κ_*$ a term characterizing the magnitude of the expected reward non-linearity. Our result is tight in its dependence on $T$, $γ_T$, and $κ_*$ for both KBs and GLBs. Finally, we present a tractable version GKB-UCB, Trac-GKB-UCB, which attains similar regret guarantees, and we discuss its time and space complexity.
format Preprint
id arxiv_https___arxiv_org_abs_2508_01681
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Generalized Kernelized Bandits: A Novel Self-Normalized Bernstein-Like Dimension-Free Inequality and Regret Bounds
Metelli, Alberto Maria
Drago, Simone
Mussi, Marco
Machine Learning
We study the regret minimization problem in the novel setting of generalized kernelized bandits (GKBs), where we optimize an unknown function $f^*$ belonging to a reproducing kernel Hilbert space (RKHS) having access to samples generated by an exponential family (EF) reward model whose mean is a non-linear function $μ(f^*)$. This setting extends both kernelized bandits (KBs) and generalized linear bandits (GLBs), providing a unified view of both settings. We propose an optimistic regret minimization algorithm, GKB-UCB, and we explain why existing self-normalized concentration inequalities used for KBs and GLBs do not allow to provide tight regret guarantees. For this reason, we devise a novel self-normalized Bernstein-like dimension-free inequality that applies to a Hilbert space of functions with bounded norm, representing a contribution of independent interest. Based on it, we analyze GKB-UCB, deriving a regret bound of order $\widetilde{O}( γ_T \sqrt{T/κ_*})$, being $T$ the learning horizon, $γ_T$ the maximal information gain, and $κ_*$ a term characterizing the magnitude of the expected reward non-linearity. Our result is tight in its dependence on $T$, $γ_T$, and $κ_*$ for both KBs and GLBs. Finally, we present a tractable version GKB-UCB, Trac-GKB-UCB, which attains similar regret guarantees, and we discuss its time and space complexity.
title Generalized Kernelized Bandits: A Novel Self-Normalized Bernstein-Like Dimension-Free Inequality and Regret Bounds
topic Machine Learning
url https://arxiv.org/abs/2508.01681