Research on the recommendation system of Gobang game based on machine learning

Gobang is an ancient game of pure strategy for two players, which not only makes people relax physically and mentally, but also exercises thinking ability. Although the rules of Gobang are simple, its playing method is complex. If you want to win, you need to be careful step by step, which is a challenging test of the thinking ability of chess players. In many cases, amateur chess players cannot figure out the right positions to place chess pieces in a short period of time, and at the same time, chess players also need to have opponents of their own level to train themselves. Therefore, a Gobang chess recommendation system is very necessary. Existing methods of Gobang recommendation mainly focus on rule-based manners, which is inaccurate and inefficient. Hence, in this paper, a Gobang recommendation system based on machine learning is proposed. We provide the details of our proposed method, and evaluate it using real-world Gobang playing scenarios. The experimental results show that our proposed machine learning methods highly improves on the existing rule-based mechanism and the random forest based method achieves the best accuracy of 93.39%.