Exploring Lunarlander V2
Let's dive into the details surrounding Lunarlander V2.
- Solving the Gym environment
- Reinforcement Learning (vanilla policy gradients) to land on the Moon.
- 3 layer Neural Network that learned to land space craft on surface with reduced fuel consumption through reinforcement learning ...
- DQN Baseline hyperaparameters. DQN hyperparameters tuned by CMA-ES.
- Showcasing an agent trained with DQN on the
In-Depth Information on Lunarlander V2
Timecodes: 0:00 Introduction 0:25 Code overview 1:08 Solution explanation 5:52 Conclusion. Solved with Deep Reinforcement Learning (DQN) in 1000 episodes. Solving OpenAI's This is a video recording the progression of my Deep Q-Network Agent on the
Training
That wraps up our extensive overview of Lunarlander V2.
