An implementation of the linear programming approach for Q-learning.
For more details on the LP approach for Q-learning check out our paper https://arxiv.org/abs/2003.08721
In order to install this python package, execute the following steps:
- create a conda environment with python 3.6 with the command
conda create -n yourenvname python=3.6 anaconda
- activate your virtual environment with
source activate yourenvname
- install gurobi with the command:
conda install gurobi
- request an academic licence for Gurobi (see https://www.gurobi.com/academia/academic-program-and-licenses/)
- clone the repo
- cd into the cloned repo on your local machine
- run the command:
python setup.py install
To check the installation, go to the folder \files
and run the command: sh bash.sh
.
Note: if you make use of this toolbox for your research, please cite https://arxiv.org/abs/2003.08721