Projects from the DeepRL Nanodegree by Udacity

This repo contains implementations to the projects of the Deep Reinforcement Learning Nanodegree by Udacity. The projects we have to implement as part of the course are the following:

• Simple navigation using DQN: Implement a DQN-based agent to collect bananas in a virtual environment.
• Continuous control with DDPG and PPO: Implement a (DDPG|PPO)-based agent to control its arm to reach a target location.
• Cooperation (Open ended project): Train two agents in a cooperative way to play tennis.

Project 1: DQN

CODE | POSTS: FULL - PART-1 - PART-2 - PART-3 | REPORT

Project 2: DDPG

CODE | REPORT

Project 3: MADDPG

CODE | REPORT