Single- versus Multiobjective Optimization for Evolution of Neural Controllers in Ms. Pac-Man
The objective of this study is to focus on the automatic generation of game artificial intelligence (AI) controllers for Ms. Pac-Man agent by using artificial neural network (ANN) and multiobjective artificial evolution. The Pareto Archived Evolution Strategy (PAES) is used to generate a Pareto opti...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2013-01-01
|
| Series: | International Journal of Computer Games Technology |
| Online Access: | http://dx.doi.org/10.1155/2013/170914 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832564348884615168 |
|---|---|
| author | Tse Guan Tan Jason Teo Kim On Chin |
| author_facet | Tse Guan Tan Jason Teo Kim On Chin |
| author_sort | Tse Guan Tan |
| collection | DOAJ |
| description | The objective of this study is to focus on the automatic generation of game artificial intelligence (AI) controllers for Ms. Pac-Man agent by using artificial neural network (ANN) and multiobjective artificial evolution. The Pareto Archived Evolution Strategy (PAES) is used to generate a Pareto optimal set of ANNs that optimize the conflicting objectives of maximizing Ms. Pac-Man scores (screen-capture mode) and minimizing neural network complexity. This proposed algorithm is called Pareto Archived Evolution Strategy Neural Network or PAESNet. Three different architectures of PAESNet were investigated, namely, PAESNet with fixed number of hidden neurons (PAESNet_F), PAESNet with varied number of hidden neurons (PAESNet_V), and the PAESNet with multiobjective techniques (PAESNet_M). A comparison between the single- versus multiobjective optimization is conducted in both training and testing processes. In general, therefore, it seems that PAESNet_F yielded better results in training phase. But the PAESNet_M successfully reduces the runtime operation and complexity of ANN by minimizing the number of hidden neurons needed in hidden layer and also it provides better generalization capability for controlling the game agent in a nondeterministic and dynamic environment. |
| format | Article |
| id | doaj-art-9e717f87c59c4cd9a331637ba5b1685d |
| institution | Kabale University |
| issn | 1687-7047 1687-7055 |
| language | English |
| publishDate | 2013-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | International Journal of Computer Games Technology |
| spelling | doaj-art-9e717f87c59c4cd9a331637ba5b1685d2025-02-03T01:11:19ZengWileyInternational Journal of Computer Games Technology1687-70471687-70552013-01-01201310.1155/2013/170914170914Single- versus Multiobjective Optimization for Evolution of Neural Controllers in Ms. Pac-ManTse Guan Tan0Jason Teo1Kim On Chin2Evolutionary Computing Laboratory, School of Engineering and Information Technology, Universiti Malaysia, Jalan (UMS), 88400 Kota Kinabalu, Sabah, MalaysiaEvolutionary Computing Laboratory, School of Engineering and Information Technology, Universiti Malaysia, Jalan (UMS), 88400 Kota Kinabalu, Sabah, MalaysiaEvolutionary Computing Laboratory, School of Engineering and Information Technology, Universiti Malaysia, Jalan (UMS), 88400 Kota Kinabalu, Sabah, MalaysiaThe objective of this study is to focus on the automatic generation of game artificial intelligence (AI) controllers for Ms. Pac-Man agent by using artificial neural network (ANN) and multiobjective artificial evolution. The Pareto Archived Evolution Strategy (PAES) is used to generate a Pareto optimal set of ANNs that optimize the conflicting objectives of maximizing Ms. Pac-Man scores (screen-capture mode) and minimizing neural network complexity. This proposed algorithm is called Pareto Archived Evolution Strategy Neural Network or PAESNet. Three different architectures of PAESNet were investigated, namely, PAESNet with fixed number of hidden neurons (PAESNet_F), PAESNet with varied number of hidden neurons (PAESNet_V), and the PAESNet with multiobjective techniques (PAESNet_M). A comparison between the single- versus multiobjective optimization is conducted in both training and testing processes. In general, therefore, it seems that PAESNet_F yielded better results in training phase. But the PAESNet_M successfully reduces the runtime operation and complexity of ANN by minimizing the number of hidden neurons needed in hidden layer and also it provides better generalization capability for controlling the game agent in a nondeterministic and dynamic environment.http://dx.doi.org/10.1155/2013/170914 |
| spellingShingle | Tse Guan Tan Jason Teo Kim On Chin Single- versus Multiobjective Optimization for Evolution of Neural Controllers in Ms. Pac-Man International Journal of Computer Games Technology |
| title | Single- versus Multiobjective Optimization for Evolution of Neural Controllers in Ms. Pac-Man |
| title_full | Single- versus Multiobjective Optimization for Evolution of Neural Controllers in Ms. Pac-Man |
| title_fullStr | Single- versus Multiobjective Optimization for Evolution of Neural Controllers in Ms. Pac-Man |
| title_full_unstemmed | Single- versus Multiobjective Optimization for Evolution of Neural Controllers in Ms. Pac-Man |
| title_short | Single- versus Multiobjective Optimization for Evolution of Neural Controllers in Ms. Pac-Man |
| title_sort | single versus multiobjective optimization for evolution of neural controllers in ms pac man |
| url | http://dx.doi.org/10.1155/2013/170914 |
| work_keys_str_mv | AT tseguantan singleversusmultiobjectiveoptimizationforevolutionofneuralcontrollersinmspacman AT jasonteo singleversusmultiobjectiveoptimizationforevolutionofneuralcontrollersinmspacman AT kimonchin singleversusmultiobjectiveoptimizationforevolutionofneuralcontrollersinmspacman |