A Data-Efficient Method of Deep Reinforcement Learning for Chinese Chess

Abstract

The computer game is the Drosophila in the field of artificial intelligence. Recently, a series of computer game systems., such as AlphaGo and AlphaGo Zero, defeating the world human champion of Go, has greatly refreshed people's understanding of the creativity of machine. This paper applies the deep reinforcement learning method to the computer Chinese Chess. We are committed to decrease the demand for computing resources heavily from multi-perspectives, such as data augmentation and using more intermediate results as labels. The experiment shows that the level of our program is increased rapidly.