Aim & Scope

Reinforcement learning is an area of machine learning seeking to provide a computational approach to understanding and automating goal-directed learning and decision-making. It addresses the question of how an autonomous agent that senses and acts in its environment can learn to choose optimal actions to achieve its goals. The approach originates from previous work in psychology (particularly animal learning), computer science (particularly dynamic programming), with ongoing work in artificial intelligence (particularly stochastic, symbolic and connectionist learning). More recently, reinforcement learning has been used to provide cognitive models that simulate human performance during problem solving and/or skill acquisition.

This thematic issue of the Journal of Algorithms seeks to celebrate the increasingly multidisciplinary nature of reinforcement learning and, in line with the Journal's manifesto, it proposes to study and present the subject from an algorithmic perspective that we refer to as Algorithmic Reinforcement Learning (ARL). It is hoped in this way that this thematic volume will serve as a reference in the area, and will help organise and promote the research across sub-areas.

We welcome the submission of innovative and mature results in specifying, developing and experimenting with ARL. Approaches that relate, compare and contrast, combine or integrate different areas of reinforcement learning are particularly encouraged. Papers describing innovative developments in the area are also encouraged. Areas of interest include, but are not limited to, the following topics:

• Multi-agent reinforcement learning
• Relational reinforcement learning
• Neuro-symbolic reinforcement learning
• Bayesian reinforcement learning
• Reinforcement learning and logic/ILP
• Reinforcement learning with background knowledge
• Robust reinforcement learning
• Reinforcement learning in game theory and bounded rationality
• Applications

Important Dates

Submission Deadline: 1st October 2008

Acceptance Notice: 20th January 2009

Final Manuscript: 1st March 2009

Publication Date: 2nd Quarter, 2009 (Tentative)

Submission Guidelines

The work submitted must be in the form of high quality, original papers, which are not simultaneously submitted for publication elsewhere. Papers should be formatted according to the journal style, and not exceed 25 pages including figures, references, etc. The papers must be submitted by sending a PDF version of the complete manuscript to arl-guest-eds@cs.rhul.ac.uk.