Persistent Reinforcement Learning Environment with Real-Time Configuration

[s-p-n]

reinforce-squared is available for free (as in beer) on Github:

You must be registered to see links

Reinforce-Squared is a machine learning environment with a display window, a graph, and a console used to enter commands that alter your model or the environment. The project is based on a video tutorial by Sentdex, and that video goes great with reinforce-squared, as you can learn what some of the configurations do, and make pretty smart blue dots.

This project is only a few days old as of this writing, so please post here or create an issue (on github) if something doesn't work right.

 

[Joopie]

Interesting! Never heard of it before. Will look at it in the weekend. Thanks!