Pack Light on the Move: Exploitation and Exploration in a Dynamic Environment

Marco LiCalzi; Davide Marchiori

doi:10.1007/978-3-319-00912-4_16

Pack Light on the Move: Exploitation and Exploration in a Dynamic Environment

Research output: Chapter in Book/Report/Conference proceeding › Book chapter › Research › peer-review

Abstract

This paper revisits a recent study by Posen and Levinthal (Manag Sci 58:587–601, 2012) on the exploration/exploitation tradeoff for a multi-armed bandit problem, where the reward probabilities undergo random shocks. We show that their analysis suffers two shortcomings: it assumes that learning is based on stale evidence, and it overlooks the steady state. We let the learning rule endogenously discard stale evidence, and we perform the long run analyses. The comparative study demonstrates that some of their conclusions must be qualified.

Original language	English
Title of host publication	Artificial Economics and Self Organization : Agent-Based Approaches to Economics and Social Systems
Editors	Stephan Leitner, Friederike Wall
Publisher	Springer
Publication date	2014
Pages	205-216
ISBN (Print)	9783319009117
ISBN (Electronic)	9783319009124
DOIs	https://doi.org/10.1007/978-3-319-00912-4_16
Publication status	Published - 2014

Series	Lecture Notes in Economics and Mathematical Systems
Volume	669
ISSN	0075-8442

Documents & Links

10.1007/978-3-319-00912-4_16

Cite this

LiCalzi, M., & Marchiori , D. (2014). Pack Light on the Move: Exploitation and Exploration in a Dynamic Environment. In S. Leitner, & F. Wall (Eds.), Artificial Economics and Self Organization: Agent-Based Approaches to Economics and Social Systems (pp. 205-216). Springer. Lecture Notes in Economics and Mathematical Systems Vol. 669 https://doi.org/10.1007/978-3-319-00912-4_16

@inbook{517ceb92d497411d9793d104534faf2d,

title = "Pack Light on the Move: Exploitation and Exploration in a Dynamic Environment",

abstract = "This paper revisits a recent study by Posen and Levinthal (Manag Sci 58:587–601, 2012) on the exploration/exploitation tradeoff for a multi-armed bandit problem, where the reward probabilities undergo random shocks. We show that their analysis suffers two shortcomings: it assumes that learning is based on stale evidence, and it overlooks the steady state. We let the learning rule endogenously discard stale evidence, and we perform the long run analyses. The comparative study demonstrates that some of their conclusions must be qualified.",

author = "Marco LiCalzi and Davide Marchiori",

year = "2014",

doi = "10.1007/978-3-319-00912-4_16",

language = "English",

isbn = "9783319009117",

series = "Lecture Notes in Economics and Mathematical Systems",

publisher = "Springer",

pages = "205--216",

editor = "Stephan Leitner and Friederike Wall",

booktitle = "Artificial Economics and Self Organization",

address = "Germany",

}

LiCalzi, M & Marchiori , D 2014, Pack Light on the Move: Exploitation and Exploration in a Dynamic Environment. in S Leitner & F Wall (eds), Artificial Economics and Self Organization: Agent-Based Approaches to Economics and Social Systems. Springer, Lecture Notes in Economics and Mathematical Systems, vol. 669, pp. 205-216. https://doi.org/10.1007/978-3-319-00912-4_16

Pack Light on the Move : Exploitation and Exploration in a Dynamic Environment. / LiCalzi, Marco; Marchiori , Davide.

Artificial Economics and Self Organization: Agent-Based Approaches to Economics and Social Systems. ed. / Stephan Leitner; Friederike Wall. Springer, 2014. p. 205-216 (Lecture Notes in Economics and Mathematical Systems, Vol. 669).

Research output: Chapter in Book/Report/Conference proceeding › Book chapter › Research › peer-review

TY - CHAP

T1 - Pack Light on the Move

T2 - Exploitation and Exploration in a Dynamic Environment

AU - LiCalzi, Marco

AU - Marchiori , Davide

PY - 2014

Y1 - 2014

N2 - This paper revisits a recent study by Posen and Levinthal (Manag Sci 58:587–601, 2012) on the exploration/exploitation tradeoff for a multi-armed bandit problem, where the reward probabilities undergo random shocks. We show that their analysis suffers two shortcomings: it assumes that learning is based on stale evidence, and it overlooks the steady state. We let the learning rule endogenously discard stale evidence, and we perform the long run analyses. The comparative study demonstrates that some of their conclusions must be qualified.

AB - This paper revisits a recent study by Posen and Levinthal (Manag Sci 58:587–601, 2012) on the exploration/exploitation tradeoff for a multi-armed bandit problem, where the reward probabilities undergo random shocks. We show that their analysis suffers two shortcomings: it assumes that learning is based on stale evidence, and it overlooks the steady state. We let the learning rule endogenously discard stale evidence, and we perform the long run analyses. The comparative study demonstrates that some of their conclusions must be qualified.

U2 - 10.1007/978-3-319-00912-4_16

DO - 10.1007/978-3-319-00912-4_16

M3 - Book chapter

SN - 9783319009117

T3 - Lecture Notes in Economics and Mathematical Systems

SP - 205

EP - 216

BT - Artificial Economics and Self Organization

A2 - Leitner, Stephan

A2 - Wall, Friederike

PB - Springer

ER -

LiCalzi M, Marchiori D. Pack Light on the Move: Exploitation and Exploration in a Dynamic Environment. In Leitner S, Wall F, editors, Artificial Economics and Self Organization: Agent-Based Approaches to Economics and Social Systems. Springer. 2014. p. 205-216. (Lecture Notes in Economics and Mathematical Systems, Vol. 669). https://doi.org/10.1007/978-3-319-00912-4_16

Pack Light on the Move: Exploitation and Exploration in a Dynamic Environment

Abstract

Documents & Links

Fingerprint

Cite this