Real-time Strategy Game Tactical Recommendation Based on Bayesian Network
Yang Zhena, Wanpeng Zhang, Liu Hongfu · Journal of Physics Conference Series · 2019
Real-time strategy (RTS) games simulate battles between large numbers of units and pose significant challenges to artificial intelligence which are complex areas of confrontation. In the field of RTS games, confrontation planning under uncertainty remains an unsolved challenge. There are two main types of uncertainty in RTS games. The first is the partial observability of the game, causing uncertainty. Second, there is uncertainty as the game is against the game and the player cannot predict what the opponent will do. This paper uses the Bayesian model as a logical alternative to solve the uncertainty caused by the confrontation game. The experimental results show that the proposed model can effectively solve the tactical recommendation problem in real-time strategy games.