IEEE Transactions on Computational Intelligence and AI in Games
Special Issue on Deep/Reinforcement Learning and Games
Guest Editors: I-Chen Wu, Chang-Shing Lee, Yuandong Tian, and Martin Müller
Submission Deadline: April 15, 2017
Deep Learning (DL) and Reinforcement Learning (RL) have been applied with
great success to many games, including Go and Atari 2600 games. Monte Carlo
Tree Search (MCTS), developed in 2006, can be viewed as a kind of online
RL. This technique has greatly improved the level of Go-playing programs.
MCTS has since become the state of the art for many other games including
Hex, Havannah, and General Game Playing, and has found much success in
applications as diverse as scheduling, unit commitment problems, and
Deep learning has transformed fields such as image and video recognition
and speech understanding. In computer games, DL started making its mark in
2014, when teams from University of Edinburgh and Google DeepMind
independently applied Deep Convolutional Neural Networks (DCNNs) to the
problem of expert move prediction in Go. Clark and Storkey's DCNN achieved
a move prediction rate of 44%, exceeding all previously published results.
DeepMind’s publication followed soon after, with a DCNN that reached 55%.
The combination of DL and RL led to great advances in Atari 2600 game
playing, and to the ultimate breakthrough in computer Go. What is the
larger impact of these new techniques? For which games do they succeed or
fail? How can they be extended to new applications? How can they be
combined with other approaches? The purpose of this special issue is to
publish high quality papers reporting the latest research covering the
theory and practice of DL/RL/DRL methods applied to games. Topics include
but are not limited to:
* MCTS and reinforcement learning
* Deep/reinforcement learning for all kinds of games, including board
games, card games, video games, general game playing, etc.
* Deep/reinforcement learning for procedural content generation (PCG)
* Deep/reinforcement learning for modeling players/designers
* Deep/reinforcement learning for game analytics
* Deep learning neural net architectures
* Online and offline deep reinforcement learning methods
* Training and testing issues for deep learning, such as transfer learning,
dropout, regularization to avoid overfitting, adaptive learning rates,
momentum, selection of training data sets, etc.
* Approximation methods for deep learning
* Hybrid deep learning approaches
* Real world applications
* DL-based knowledge representation models for games
Authors should follow normal TCIAIG guidelines for their submissions, but
clearly identify their papers for this special issue during the submission
process. See http://www.ieee-cis.org/pubs/tciaig/ for author information.
Extended versions of previously published conference papers are welcome,
provided that the journal paper is a significant extension, and is
accompanied by a cover letter explaining the additional contribution. Short
papers or correspondences describing novel experimental results are also
welcome. The deadline for submissions is April 15, 2017.
Computer-go mailing list