A UAV Pursuit-Evasion Strategy Based on DDPG and Imitation Learning

Table 1

Table of training hyperparameter.

Training hyperparameter	Symbol	Value
Discounting factor		0.9
Inertial update rate		0.01
Memory size		30000
Size of batch experience	Batch size	64
Simulation time step		0.1
Learning rate of Critic network		0.002
Learning rate of Actor network		0.001
Number of episodes	MaxEpisode	4000
Number of steps in one episode	MaxStep	300