DOI: 10.5176/978-981-08-8227-3_cgat08-45
Authors: Arsa Tangchitsomkit and Supot Nitsuwat
Abstract:
The research aimed to develop the artificial intelligence system for fighting games by using Length Learning Techniques. The results of the research were found that (1) data classification and decision tree
theory were used in Learning System as they could produce flexible parameter so that each character could express its behavior appropriately. (2) Fuzzy Logic System which represented with degrees of imprecision was used for decision making for appropriately character’s behavior and being consistent to the input data. (3) Length Judge Line System was used to check the level of attacking objects. Input crisp value was converted into fussy set and created into length judge line table in order to weigh the parameter value for the attacking objects value. (4) Finite state machine theory was used to express each character’s behavior. When the outputs of Learning System or Fuzzy Logic System and Length Judge Line System were found, the character would be assigned to express its behavior. When Length Learning Technique was applied to use in fighting games, it was easy to set for the opponent’s attack distance. The character expresses its behavior to protect itself in the safe distance. Parameter was also flexibly adjusted so that the character could express its behavior appropriately.
Updating...