Actor-Critic Methods

Back to Home

01. Introduction
02. Motivation
03. Bias and Variance
04. Two Ways for Estimating Expected Returns
05. Baselines and Critics
06. Policy-based, Value-Based, and Actor-Critic
07. A Basic Actor-Critic Agent
08. A3C: Asynchronous Advantage Actor-Critic, N-step
09. A3C: Asynchronous Advantage Actor-Critic, Parallel Training
10. A3C: Asynchronous Advantage Actor-Critic, Off- vs On-policy
11. A2C: Advantage Actor-Critic
12. A2C Code Walk-through
13. GAE: Generalized Advantage Estimation
14. DDPG: Deep Deterministic Policy Gradient, Continuous Actions
15. DDPG: Deep Deterministic Policy Gradient, Soft Updates
16. DDPG Code Walk-through
17. Summary

Back to Home

13. GAE: Generalized Advantage Estimation

GAE: Generalized Advantage Estimation

M3 L5 13 GAE Generalized Advantage Estimation V2

Link to the GAE paper: https://arxiv.org/abs/1506.02438

udacimak v1.2.1