Description

This pull request implements the DDPG (Deep Deterministic Policy Gradient) algorithm, along with 2 test cases.

Implementation details

DDPG is an actor-critic algorithm designed for continuous action spaces. It combines deep neural networks with deterministic policy gradients to learn optimal policies in a continuous control setting.

Implemented four networks:

• policyNetwork (actor network)
• targetPNetwork (target actor network)
• learningQNetwork (critic network)
• targetQNetwork (target critic network)

How Has This Been Tested?

• Included a Pendulum test that successfully passes on my machine. The average reward achieved in the Pendulum environment is approximately -500, indicating successful learning.
• Additionally, added a test for continuous action spaces, which also passes.

The test configurations for DDPG are adapted from the SAC (Soft Actor-Critic) implementation since both DDPG and SAC are policy gradient off-policy algorithms. This ensures consistent evaluation and comparison of the algorithms.