[Retracted] An Efficient and Reliable Routing Method for Hybrid Mobile Ad Hoc Networks Using Deep Reinforcement Learning

Figure 10

Processing the output of the U-net-similar part to achieve dueling Q-learning.