Reinforcement Learning using texplore package

asked 2014-03-25 18:52:42 -0600

dnth gravatar image

I have been using the texplore package for learning. As you run the algorithm, at what point does the algorithm stop learning? For example

I ran th below command for the environment

rosrun rl_env env --env taxi --stochastic

and below command for the agent

rosrun rl_agent agent --agent sarsa

But running a while, I noticed that the algorithm doesnt stop learning. It would go on for thousands of episode! How do I go about this?

edit retag flag offensive close merge delete

Comments

Dear dnth, I myself need to use texplore package to implement it on some simple robot in gazebosim. Please guide how should I start. It would be highly appreciated if you can send me your email address as well as I hope to be using texplore (rather learning to use it) for next 5 to 6 months! My email address: mctncontrol@gmail.com

mudassar gravatar image mudassar  ( 2014-03-29 07:31:18 -0600 )edit

Hi! You can read the guide on the tutorial page which is sufficient in understanding the package. My email is dickson.neoh AT gmail DOT com

dnth gravatar image dnth  ( 2014-03-30 23:36:35 -0600 )edit