Mass Mahjong Decision System Based on Transfer Learning

Abstract

In this paper, we propose a transfer learning to solve the problem of lacking in data and the difficulty in constructing models effectively, which is typically represented by Mass Mahjong in the field of imperfect information. Design and implement the Mass Mahjong Discard model based on transfer learning. The previously well-trained Blood Mahjong Discard model on a large dataset is migrated to Mass Mahjong Discard model in a similar domain. In the subsequent model optimization, a self-play based approaching is used to improve the Mass Mahjong Discard model. The experimental results show that the transfer learning-based Mass Mahjong Discard model performs well in the situation of less data, and can fit the Mass Mahjong Discard rule. And the model won the second prize in the Mass Mahjong event of the National University Computer Gaming Competition in 2021.