DDQN¶
-
class
DDQN
(model, gamma=None, lr=None)[source]¶ Bases:
parl.core.paddle.algorithm.Algorithm
-
__init__
(model, gamma=None, lr=None)[source]¶ DDQN algorithm
Parameters: - model (parl.Model) – forward neural network representing the Q function.
- gamma (float) – discounted factor for accumulative reward computation
- lr (float) – learning rate.
-