contextual: Evaluating contextual multi-armed bandit problems in R

Research output: Working paperOther research output

3 Downloads (Pure)

Abstract

Over the past decade, contextual bandit algorithms have been gaining in popularity due to their effectiveness and flexibility in solving sequential decision problems---from online advertising and finance to clinical trial design and personalized medicine. At the same time, there are, as of yet, surprisingly few options that enable researchers and practitioners to simulate and compare the wealth of new and existing bandit algorithms in a standardized way. To help close this gap between analytical research and empirical evaluation the current paper introduces the object-oriented R package "contextual": a user-friendly and, through its object-oriented structure, easily extensible framework that facilitates parallelized comparison of contextual and context-free bandit policies through both simulation and offline analysis.
Original languageEnglish
PublisherarXiv.org
Number of pages55
Publication statusPublished - 6 Nov 2018

Publication series

NamearXiv

Fingerprint

Finance
Medicine
Marketing

Keywords

  • cs.LG
  • math.OC
  • stat.ML
  • 93E35
  • I.2.6; K.4.4; F.2.0

Cite this

@techreport{3651f1e634c74f2cae0a0598146990e3,
title = "contextual: Evaluating contextual multi-armed bandit problems in R",
abstract = "Over the past decade, contextual bandit algorithms have been gaining in popularity due to their effectiveness and flexibility in solving sequential decision problems---from online advertising and finance to clinical trial design and personalized medicine. At the same time, there are, as of yet, surprisingly few options that enable researchers and practitioners to simulate and compare the wealth of new and existing bandit algorithms in a standardized way. To help close this gap between analytical research and empirical evaluation the current paper introduces the object-oriented R package {"}contextual{"}: a user-friendly and, through its object-oriented structure, easily extensible framework that facilitates parallelized comparison of contextual and context-free bandit policies through both simulation and offline analysis.",
keywords = "cs.LG, math.OC, stat.ML, 93E35, I.2.6; K.4.4; F.2.0",
author = "Emden, {Robin van} and Maurits Kaptein",
note = "55 pages, 12 figures",
year = "2018",
month = "11",
day = "6",
language = "English",
series = "arXiv",
publisher = "arXiv.org",
type = "WorkingPaper",
institution = "arXiv.org",

}

contextual : Evaluating contextual multi-armed bandit problems in R. / Emden, Robin van; Kaptein, Maurits.

arXiv.org, 2018. (arXiv).

Research output: Working paperOther research output

TY - UNPB

T1 - contextual

T2 - Evaluating contextual multi-armed bandit problems in R

AU - Emden, Robin van

AU - Kaptein, Maurits

N1 - 55 pages, 12 figures

PY - 2018/11/6

Y1 - 2018/11/6

N2 - Over the past decade, contextual bandit algorithms have been gaining in popularity due to their effectiveness and flexibility in solving sequential decision problems---from online advertising and finance to clinical trial design and personalized medicine. At the same time, there are, as of yet, surprisingly few options that enable researchers and practitioners to simulate and compare the wealth of new and existing bandit algorithms in a standardized way. To help close this gap between analytical research and empirical evaluation the current paper introduces the object-oriented R package "contextual": a user-friendly and, through its object-oriented structure, easily extensible framework that facilitates parallelized comparison of contextual and context-free bandit policies through both simulation and offline analysis.

AB - Over the past decade, contextual bandit algorithms have been gaining in popularity due to their effectiveness and flexibility in solving sequential decision problems---from online advertising and finance to clinical trial design and personalized medicine. At the same time, there are, as of yet, surprisingly few options that enable researchers and practitioners to simulate and compare the wealth of new and existing bandit algorithms in a standardized way. To help close this gap between analytical research and empirical evaluation the current paper introduces the object-oriented R package "contextual": a user-friendly and, through its object-oriented structure, easily extensible framework that facilitates parallelized comparison of contextual and context-free bandit policies through both simulation and offline analysis.

KW - cs.LG

KW - math.OC

KW - stat.ML

KW - 93E35

KW - I.2.6; K.4.4; F.2.0

M3 - Working paper

T3 - arXiv

BT - contextual

PB - arXiv.org

ER -