DocumentCode :
547355
Title :
Analysis of rule based look-ahead strategy using Pacman testbed
Author :
Ali, Rabia Babar ; Munir, Ali ; Farooqi, Ashfaq Hussain
Author_Institution :
Dept. of Comput. Sci., Nat. Univ. of Comput. & Emerging Sci., Islamabad, Pakistan
Volume :
3
fYear :
2011
fDate :
10-12 June 2011
Firstpage :
480
Lastpage :
483
Abstract :
Games may be used as a test-bed for the analysis of the working of any innovative idea that lies in the field of computational or artificial intelligence. There are three general categories of such games; board games (e.g. tic-tac toe, checkers, Othello), video games (e.g. Dead-end, Cellz, Flatland, Pacman, Nero) and robotics. We have implemented a test-bed of Pacman game in C# and applied various strategies to Pacman agent. The fitness of a strategy depends upon two factors; maximum survival of the Pacman player and swallowed pallets. Experimental results favors rule based look-ahead mechanism over random, greedy and rule based with greed approaches.
Keywords :
C language; computer games; greedy algorithms; neural nets; C#; Pacman game; Pacman player; Pacman testbed; artificial intelligence; board game; computational intelligence; rule based look ahead strategy; video game; Artificial intelligence; Artificial neural networks; Computers; Evolutionary computation; Games; Humans; Artificial Intelligence (AI); Artificial Neural Network (ANN); Computational Intelligence (CI); Next movement selection strategy; Pacman and Ghosts;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Science and Automation Engineering (CSAE), 2011 IEEE International Conference on
Conference_Location :
Shanghai
Print_ISBN :
978-1-4244-8727-1
Type :
conf
DOI :
10.1109/CSAE.2011.5952723
Filename :
5952723
Link To Document :
بازگشت