A Hybrid Fuzzy ANN System for Agent Adaptation in a First Person Shooter

Algorithm 7

Pseudocode for weapon selection feedback function.

weaponFeedback()

       last_impact_time = current_time

       damage_enemy = damage to enemy

       damage_self = damage to self from splash from own

weapon

       if enemy killed

      bonus = 80

end of if

feedback_value = (damage_enemy − damage_self)/

               (last_fire_time − prev_fire_time) + bonus

feedback_value = normaliseFeedback(feedback_value)

     trainPerceptron(feedback_value)

     prev_impact_time = last_impact_time

     last_impact_time = 0

end of function