Improving Maneuver Strategy in Air Combat by Alternate Freeze Games with a Deep Reinforcement Learning Algorithm

Figure 5

Evaluation of the neural network and minibatch.