Federated Reinforcement Learning-Based UAV Swarm System for Aerial Remote Sensing

Table 3

Hyperparameters and values used for learning.

Hyperparameter	Value
Actor network dimension	162562562565
Critic network dimension	162562562565
Minibatch size	5
Number of epochs	4
Learning rate	0.0003
Horizon value	20
Generalized advantage estimator	0.95
Discount factor gamma	0.99
Clipping parameter	0.2
Value function coefficient	0.5
Optimizer algorithm	Adam