Output details
11 - Computer Science and Informatics
University of Strathclyde
Output title
An evaluation of the benefits of look-ahead in Pac-Man
Type
E - Conference contribution
Name of conference/published proceedings
IEEE Symposium Computational Intelligence and Games, 2008 : CIG '09
Volume number
-
Issue number
-
First page of article
310
ISSN of proceedings
-
Year of publication
2008
URL
-
Number of additional authors
3
Additional information
<22>The paper's significance is its originality in applying deliberative reasoning, as opposed to reactive learning, to key problems in AI, using the standard benchmark of Pac-Man. It was very influential on subsequent work by many researchers, including the seminal work of Robles and Lucas (2009). This led to a sequence of papers which apply deliberative reasoning to this problem, including the very successful Monte Carlo Tree Search work of Samothrakis et al in 2011. As a direct result of this paper, it became accepted that reactive agents cannot perform well at this problem and that some deliberation is essential.
Interdisciplinary
-
Cross-referral requested
-
Research group
C - Software Systems
Citation count
3
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-
