Saved in:
Bibliographic Details
Main Authors: Gu, Yuzhou, Han, Yanjun, Qian, Jian
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.00273
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918165259223040
author Gu, Yuzhou
Han, Yanjun
Qian, Jian
author_facet Gu, Yuzhou
Han, Yanjun
Qian, Jian
contents We study the evolution of information in interactive decision making through the lens of a stochastic multi-armed bandit problem. Focusing on a fundamental example where a unique optimal arm outperforms the rest by a fixed margin, we characterize the optimal success probability and mutual information over time. Our findings reveal distinct growth phases in mutual information -- initially linear, transitioning to quadratic, and finally returning to linear -- highlighting curious behavioral differences between interactive and non-interactive environments. In particular, we show that optimal success probability and mutual information can be decoupled, where achieving optimal learning does not necessarily require maximizing information gain. These findings shed new light on the intricate interplay between information and learning in interactive decision making.
format Preprint
id arxiv_https___arxiv_org_abs_2503_00273
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Evolution of Information in Interactive Decision Making: A Case Study for Multi-Armed Bandits
Gu, Yuzhou
Han, Yanjun
Qian, Jian
Machine Learning
Information Theory
We study the evolution of information in interactive decision making through the lens of a stochastic multi-armed bandit problem. Focusing on a fundamental example where a unique optimal arm outperforms the rest by a fixed margin, we characterize the optimal success probability and mutual information over time. Our findings reveal distinct growth phases in mutual information -- initially linear, transitioning to quadratic, and finally returning to linear -- highlighting curious behavioral differences between interactive and non-interactive environments. In particular, we show that optimal success probability and mutual information can be decoupled, where achieving optimal learning does not necessarily require maximizing information gain. These findings shed new light on the intricate interplay between information and learning in interactive decision making.
title Evolution of Information in Interactive Decision Making: A Case Study for Multi-Armed Bandits
topic Machine Learning
Information Theory
url https://arxiv.org/abs/2503.00273