Everything to reproduce reinforcement learning experiments done in my Master's thesis. For now you'll have to find your way around by yourself. In time i'll try to come up with some documentation.
Python 2.7 is used. The following packages are required:
- Theano
- RL-Glue
- configobj
- (optional) IPython Parallel for running the experiments remotelly and/or in parallel.
To run the environments you will also have to set up rihardsk/predictive-rl_envs
See rlglueexp module for information on automating the experiment process.
See the experiments directory for the experiment configuration files to reproduce the results in my thesis.