mlpack  blog
Proximal Policy Optimization - Week 4

Proximal Policy Optimization - Week 4

Xiaohong J, 23 June 2019

This week, I mainly focus on Proximal Policy Optimization pull request.

  • Added Step and Episode function in PPO algorithm.
  • Added initial test case for the PPO algorithm.
  • Opened a PR for the the pendulum environment.
  • Added action prediction in the PPO algorithm.

I choose the continuous task, pendulum, to test the PPO algorithm, it seems that some code in the original code base did not support the continuous task. Next week, I will change that one.

Thanks for reading :).