Baselines for Joint-Action Reinforcement Learning of Coordination in Cooperative Multi-agent Systems

Martin Carpenter; Daniel Kudenko

doi:10.1007/978-3-540-32274-0_4

Baselines for Joint-Action Reinforcement Learning of Coordination in Cooperative Multi-agent Systems

Martin Carpenter, Daniel Kudenko

Computer Science

Research output: Contribution to conference › Paper

Abstract

A common assumption for the study of reinforcement learning of coordination is that agents can observe each other’s actions (so-called joint-action learning). We present in this paper a number of simple joint-action learning algorithms and show that they perform very well when compared against more complex approaches such as OAL [1], while still maintaining convergence guarantees. Based on the empirical results, we argue that these simple algorithms should be used as baselines for any future research on joint-action learning of coordination.

Original language	Undefined/Unknown
Pages	55-72
DOIs	https://doi.org/10.1007/978-3-540-32274-0_4
Publication status	Published - 2005

Access to Document

10.1007/978-3-540-32274-0_4

http://springerlink.metapress.com/openurl.asp?genre=articleissn=0302-9743volume=3394spage=55

Cite this

@conference{bc270d14090341068af0677066f1fbe4,

title = "Baselines for Joint-Action Reinforcement Learning of Coordination in Cooperative Multi-agent Systems",

abstract = "A common assumption for the study of reinforcement learning of coordination is that agents can observe each other{\textquoteright}s actions (so-called joint-action learning). We present in this paper a number of simple joint-action learning algorithms and show that they perform very well when compared against more complex approaches such as OAL [1], while still maintaining convergence guarantees. Based on the empirical results, we argue that these simple algorithms should be used as baselines for any future research on joint-action learning of coordination. ",

author = "Martin Carpenter and Daniel Kudenko",

year = "2005",

doi = "10.1007/978-3-540-32274-0_4",

language = "Undefined/Unknown",

pages = "55--72",

}

TY - CONF

T1 - Baselines for Joint-Action Reinforcement Learning of Coordination in Cooperative Multi-agent Systems

AU - Carpenter, Martin

AU - Kudenko, Daniel

PY - 2005

Y1 - 2005

N2 - A common assumption for the study of reinforcement learning of coordination is that agents can observe each other’s actions (so-called joint-action learning). We present in this paper a number of simple joint-action learning algorithms and show that they perform very well when compared against more complex approaches such as OAL [1], while still maintaining convergence guarantees. Based on the empirical results, we argue that these simple algorithms should be used as baselines for any future research on joint-action learning of coordination.

AB - A common assumption for the study of reinforcement learning of coordination is that agents can observe each other’s actions (so-called joint-action learning). We present in this paper a number of simple joint-action learning algorithms and show that they perform very well when compared against more complex approaches such as OAL [1], while still maintaining convergence guarantees. Based on the empirical results, we argue that these simple algorithms should be used as baselines for any future research on joint-action learning of coordination.

U2 - 10.1007/978-3-540-32274-0_4

DO - 10.1007/978-3-540-32274-0_4

M3 - Paper

SP - 55

EP - 72

ER -