Digging in (month 3)

The goal for this month was to set the whole environment to work and start playing with reinforcement learning so the lectures can be consolidated with some exercises.

Lectures

• Thesis of a DQN neural network able to beat an human gamer playing ATARI games

• Study of the effect of convolutional masks making use of the Fourier Transform

• Reinforcement learning introduction

• DDPG introduction

• DQN introduction and implementation explanation video

• DQN hyperparameters tunning using openAI cartpole exercise

Lab work

• OpenAi gym installation and the acomplishment of the exercises explained in Shiva Verma’s blog