Road to an Interesting Opponent : An Agent that Predicts the Users ' Combination Attacks in a Fighting Videogame

In fighting videogames users usually prefer playing against other users rather than against the machine. We are assuming that the adaptability of the human player makes it interesting. We are aiming to produce an agent for a fighting videogame that can adapt to its users, allowing users to enjoy the game even when playing alone. We have completed a part of this continuing project: an agent that predicts the users' combination attacks. We introduce this agent and present the results of the experiments.

[1]  Thore Graepel,et al.  LEARNING TO FIGHT , 2004 .

[2]  Georgios N. Yannakakis,et al.  Evolving opponents for interesting interactive computer games , 2004 .

[3]  Nikos Mamoulis,et al.  Mining frequent spatio-temporal sequential patterns , 2005, Fifth IEEE International Conference on Data Mining (ICDM'05).

[4]  Ernest Adams,et al.  Fundamentals of Game Design , 2006 .

[5]  Richard S. Sutton,et al.  Reinforcement Learning: An Introduction , 1998, IEEE Trans. Neural Networks.