Collaborative Multi-Agent Bandits

Junghyun Lee
Oct 27, 2020 project

(Part of Samsung & IITP RL project)

Project #1. Flooding with Absorption: An Efficient Protocol for Heterogeneous Bandits over Complex Networks

Accepted to OPODIS 2023. Joint work with Laura Schmid and Se-Young Yun (KAIST AI).

Bandits RL Networks
Junghyun Lee
Junghyun Lee
PhD Student

PhD student at GSAI, KAIST, jointly advised by Prof. Se-Young Yun and Prof. Chulhee Yun. Interested in mathematical and theoretical AI, where I am interested in explaining the recent success of machine/deep learning algorithms and designing mathematically well-grounded algorithms for various scenarios.

Posts

Our OPODIS 2023 paper has received best student paper award!!
(to be filled out…)
Junghyun Lee
Nov 21, 2023 1 min read
Project
Our preprint on improved confidence set construction for (multinomial) logistic bandits now available on arXiv!
(to be filled out…)
Junghyun Lee
Oct 31, 2023 1 min read
Project
One paper accepted to OPODIS 2023!
(to be filled out…)
Junghyun Lee
Oct 30, 2023 1 min read
Project
Our arXiv paper on collaborative bandits over a network has been updated!
(to be filled out…)
Junghyun Lee
Sep 22, 2023 1 min read
Project
Our preprint on multi-agent heterogeneous bandits over a communication network now available on arXiv!
(to be filled out…)
Junghyun Lee
Mar 10, 2023 1 min read
Project

Publications

Flooding with Absorption: An Efficient Protocol for Heterogeneous Bandits over Complex Networks
A novel problem setting where heterogeneous multi-agent bandits collaborate over a network to minimize their group regret. To deal with the high communication complexity of the classic flooding protocol combined with UCB, a new network protocol called Flooding with Absorption (FwA) is proposed. Theoretical and empirical analyses are provded for flooding and FwA, showing the efficacy of our proposed FwA.
Junghyun Lee, Laura Schmid, Se-Young Yun
PDF Cite Code Project Poster Slides
Flooding with Absorption: An Efficient Protocol for Heterogeneous Bandits over Complex Networks