[ros-users] [Discourse.ros.org] [general] A toolkit for Reinforcement Learning using ROS and Gazebo

[Ross Story]

Sure, here are some references and reading material.

A benchmark that sadly doesn't include DQN, but does include TRPO and DDPG:
https://arxiv.org/pdf/1604.06778v3.pdf

DDPG:
https://arxiv.org/pdf/1509.02971v5.pdf

Asynchronous RL learning showing improved performance of asynchronous actor critic over asynchronous Q learning.
https://arxiv.org/pdf/1602.01783v2.pdf

@spk921 Apologies, it's NAF not FAN. It was designed for robotic manipulation and outperforms DDPG. Paper is here:
https://arxiv.org/pdf/1610.00633v1.pdf






---
[Visit Topic](https://discourse.ros.org/t/a-toolkit-for-reinforcement-learning-using-ros-and-gazebo/442/14) or reply to this email to respond.