List view

• More Policy Gradients

  - Read more into policy gradients; e.g. TRPO paper - Perhaps do the Berkeley Homework on implementing basic policy gradients. - Implement more advanced versions of policy gradients?

  Overdue by 5 year(s)

  •

  Due by August 5, 2020

  •6/8 issues closed
  75% complete2 open6 closed

• Delve into Deep Q-Learning and Inverse RL

  Learn the SOTA methods for DQN - Dueling DQN - Double DQN - DQfD (Deep Q Learning from Demonstrations) Learn about Inverse RL and how we can use it as a replacement for manual subrewards from ForgER

  Overdue by 5 year(s)

  •

  Due by August 19, 2020

  •3/7 issues closed
  42% complete4 open3 closed