A game AI based on ID3 algorithm

Abstract

Traditional game AI is aimed at the standard players in the beginning of the construction. After designing the AI, The whole game decision-making system will not be dynamically adjusted, Because of the strong regularity, the traditional game AI is easy to be found the weakness by users, which is not easy to maintain the rationality of the decision. This paper presents a design scheme of game AI based on decision tree algorithm ID3. It through the mining of the player with the NPC interactive data on the set target has a significant correlation of data clues, real-time control of the decision-making system, to maintain the rationality of the decision. The experimental results show that the proposed scheme can adapt to the player's operation well, and it can keep the rationality of the decision in the actual operating environment.