clement-chupin/BenchNeuralNework

Evaluate Feature extraction in the reinforcement learning, with environements from OpenAi gym, and policies from stable-baselines3

In this project, we evaluate feature extraction in the RL filed

report : https://github.com/clement-chupin/BenchNeuralNework/blob/master/documentation/report/Rapport_projet_M2_CHUPIN_Clement.pdf

Feature extraction by activation function

We will project our input into activation function, and evaluate the gains for our RL agent

Result of the feature extraction

Finaly, we can evaluate almost everything in terms of Feature Extraction

To show the result, we can plot environements and policies on the same grid