Empirical methods for artificial intelligence
Empirical methods for artificial intelligence
Introduction to Reinforcement Learning
Introduction to Reinforcement Learning
Artificial Intelligence: A Modern Approach
Artificial Intelligence: A Modern Approach
Toward an understanding of flow in video games
Computers in Entertainment (CIE) - Theoretical and Practical Computer Applications in Entertainment
| Hi-index | 0.01 |
In computer games, tutoring systems are used for two purposes: (1) to introduce a human player to the mechanics of a game, and (2) to ensure that the computer plays the game at a level of playing strength that is appropriate for the skills of a novice human player. Regarding the second purpose, the issue is not to produce occasionally a weak move (i.e., a give-away move) so that the human player can win, but rather to produce not-so-strong moves under the proviso that, on a balance of probabilities, they should go unnoticed. This paper focuses on using adaptive game AI to implement a tutoring system for commercial games. We depart from the novel learning technique ‘dynamic scripting’ and add three straightforward enhancements to achieve an ‘even game’, viz. high-fitness penalising, weight clipping, and top culling. Experimental results indicate that top culling is particularly successful in creating an even game. Hence, our conclusion is that dynamic scripting with top culling can implement a successful tutoring system for commercial games.