Introduction of Reinforcement Learning from OpenAI gym library
Created homemade Framework (only DQN for the moment) in order to :
- Master every mathematical aspects of the algorithm
- Garantee that Agents are compatible to any environment (DQN will later be used to create a Coinbase Bot)
Results on Homemade DQN using Cartpole Environment in gym (Maximum Reward 200):
