Deep Reinforcement Learning
Module 2

Policy Gradient Algorithms

VPG, TRPO, and PPO — the on-policy family of algorithms that directly optimize the policy objective.

3 readings 1 quiz 1 lab
Readings
1
Vanilla Policy Gradient (VPG)
12 min
2
Trust Region Policy Optimization (TRPO)
12 min
3
Proximal Policy Optimization (PPO)
12 min
Quizzes
Q
On-Policy Algorithms Quiz
6 questions · 70% to pass
Labs
VPG & PPO in PyTorch
60 min