Deep Deterministic Policy Gradient (DDPG)

Target Networks

Duplicated neural networks with slowly updated weights to stabilize learning by providing more consistent targets.

← Zurück