Hybrid Online and Offline Reinforcement Learning for Tibetan Jiu Chess

Figure 4

The deep neural network structure.