High-dimensional Nonparametric Contextual Bandit Problem

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Iwazaki, Shogo, Komiyama, Junpei, Imaizumi, Masaaki
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916746091298816
author Iwazaki, Shogo
Komiyama, Junpei
Imaizumi, Masaaki
author_facet Iwazaki, Shogo
Komiyama, Junpei
Imaizumi, Masaaki
contents We consider the kernelized contextual bandit problem with a large feature space. This problem involves $K$ arms, and the goal of the forecaster is to maximize the cumulative rewards through learning the relationship between the contexts and the rewards. It serves as a general framework for various decision-making scenarios, such as personalized online advertising and recommendation systems. Kernelized contextual bandits generalize the linear contextual bandit problem and offers a greater modeling flexibility. Existing methods, when applied to Gaussian kernels, yield a trivial bound of $O(T)$ when we consider $Ω(\log T)$ feature dimensions. To address this, we introduce stochastic assumptions on the context distribution and show that no-regret learning is achievable even when the number of dimensions grows up to the number of samples. Furthermore, we analyze lenient regret, which allows a per-round regret of at most $Δ> 0$. We derive the rate of lenient regret in terms of $Δ$.
format Preprint
id arxiv_https___arxiv_org_abs_2505_14102
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle High-dimensional Nonparametric Contextual Bandit Problem
Iwazaki, Shogo
Komiyama, Junpei
Imaizumi, Masaaki
Machine Learning
Methodology
We consider the kernelized contextual bandit problem with a large feature space. This problem involves $K$ arms, and the goal of the forecaster is to maximize the cumulative rewards through learning the relationship between the contexts and the rewards. It serves as a general framework for various decision-making scenarios, such as personalized online advertising and recommendation systems. Kernelized contextual bandits generalize the linear contextual bandit problem and offers a greater modeling flexibility. Existing methods, when applied to Gaussian kernels, yield a trivial bound of $O(T)$ when we consider $Ω(\log T)$ feature dimensions. To address this, we introduce stochastic assumptions on the context distribution and show that no-regret learning is achievable even when the number of dimensions grows up to the number of samples. Furthermore, we analyze lenient regret, which allows a per-round regret of at most $Δ> 0$. We derive the rate of lenient regret in terms of $Δ$.
title High-dimensional Nonparametric Contextual Bandit Problem
topic Machine Learning
Methodology
url https://arxiv.org/abs/2505.14102