NDLI: Arm Space Decomposition as a Strategy for Tackling Large Scale Multi-armed Bandit Problems

Content Provider	IEEE Xplore Digital Library
Author	Gupta, N. Granmo, O.-C. Agrawala, A.
Copyright Year	2013
Description	Author affiliation: Univ. of Agder, Grimstad, Norway (Granmo, O.-C.) \|\| Univ. of Maryland, College Park, MD, USA (Gupta, N.; Agrawala, A.)
Abstract	Recent multi-armed bandit based optimization schemes provide near-optimal balancing of arm exploration against arm exploitation, allowing the optimal arm to be identified with probability arbitrarily close to unity. However, the convergence speed drops dramatically as the number of bandit arms grows large, simply because singling out the optimal arm requires experimentation with all of the available arms. Furthermore, effective exploration and exploitation typically demands computational resources that grow linearly with the number of arms. Although the former problem can be remedied to some degree when prior knowledge about arm correlation is available, the latter problem persists. In this paper we propose a Thompson Sampling (TS) based scheme for exploring an arm space of size K by decomposing it into two separate arm spaces, each of size sqrtK, thus achieving sub-linear scalability. In brief, two dedicated Thompson Samplers explore each arm space separately. However, at each iteration, arm selection feedback is obtained by jointly considering the arms selected by each of the Thompson Samplers, mapping them into the original arm space. This kind of decentralized decision-making can be modeled as a game theory problem, where two independent decision makers interact in terms of a common pay-off game. Our scheme requires no communication between the decision makers, who have complete autonomy over their actions. Thus it is ideal for coordinating autonomous agents in a multi-agent system. Extensive experiments, including instances possessing multiple Nash equilibria, demonstrate remarkable performance benefits. Although TS based schemes already are among the top-performing bandit players, our proposed arm space decomposition scheme provide drastic improvements for large arm spaces, not only in terms of processing speed and memory usage, but also in terms of an improved ability to identify the optimal arm, increasing with the number of bandit arms.
Sponsorship	IEEE Syst., Man, Cybern. Soc.
Starting Page	252
Ending Page	257
File Size	450572
Page Count	6
File Format	PDF
ISBN	9780769551449
DOI	10.1109/ICMLA.2013.51
Language	English
Publisher	Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher Date	2013-12-04
Publisher Place	USA
Access Restriction	Subscribed
Rights Holder	Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subject Keyword	Game Theory Games Multi-Armed Bandits Nash equilibrium Educational institutions Bayes methods Space exploration Machine Learning Optimization
Content Type	Text
Resource Type	Article

Sl.	Authority	Responsibilities	Communication Details
1	Ministry of Education (GoI), Department of Higher Education	Sanctioning Authority	https://www.education.gov.in/ict-initiatives
2	Indian Institute of Technology Kharagpur	Host Institute of the Project: The host institute of the project is responsible for providing infrastructure support and hosting the project	https://www.iitkgp.ac.in
3	National Digital Library of India Office, Indian Institute of Technology Kharagpur	The administrative and infrastructural headquarters of the project	Dr. B. Sutradhar bsutra@ndl.gov.in
4	Project PI / Joint PI	Principal Investigator and Joint Principal Investigators of the project	Dr. B. Sutradhar bsutra@ndl.gov.in Prof. Saswat Chakrabarti will be added soon
5	Website/Portal (Helpdesk)	Queries regarding NDLI and its services	support@ndl.gov.in
6	Contents and Copyright Issues	Queries related to content curation and copyright issues	content@ndl.gov.in
7	National Digital Libarray of India Club (NDLI Club)	Queries related to NDLI Club formation, support, user awareness program, seminar/symposium, collaboration, social media, promotion, and outreach	clubsupport@ndl.gov.in
8	Digital Preservation Centre (DPC)	Assistance with digitizing and archiving copyright-free printed books	dpc@ndl.gov.in
9	IDR Setup or Support	Queries related to establishment and support of Institutional Digital Repository (IDR) and IDR workshops	idr@ndl.gov.in

Interface Design Optimization as a Multi-Armed Bandit Problem

A Multi-armed Bandit Approach to Cost-Sensitive Decision Tree Learning

Almost optimal exploration in multi-armed bandits

Bridging Computational Neuroscience and Machine Learning on Non-Stationary Multi-Armed Bandits

Playing in stochastic environment: from multi-armed bandits to two-player games

Using Multi-Arm Bandits to Optimize Game Play Metrics and Effective Game Design

Subjective games and equilibria (1994).

Large Bandit Games (2012)

Multiple identifications in multi-armed bandits.

Arm Space Decomposition as a Strategy for Tackling Large Scale Multi-armed Bandit Problems

Similar Documents

Interface Design Optimization as a Multi-Armed Bandit Problem

A Multi-armed Bandit Approach to Cost-Sensitive Decision Tree Learning

Almost optimal exploration in multi-armed bandits

Bridging Computational Neuroscience and Machine Learning on Non-Stationary Multi-Armed Bandits

Playing in stochastic environment: from multi-armed bandits to two-player games

Using Multi-Arm Bandits to Optimize Game Play Metrics and Effective Game Design

Subjective games and equilibria (1994).

Large Bandit Games (2012)

Multiple identifications in multi-armed bandits.

Arm Space Decomposition as a Strategy for Tackling Large Scale Multi-armed Bandit Problems