This is the codification used in the TiRL 2017 paper proposing the object-oriented Probabilistic Inter-Task Mapping. You are free to use all or part of the codes here presented for any purpose, provided that the paper is properly cited and the original authors properly credited. All the files here shared come with no warranties.

Paper bib entry:

@inproceedings{SilvaAndCosta2017,
author = {Silva, Felipe Leno da and
Anna Helena Reali Costa},
title = {{Towards Zero-Shot Autonomous Inter-Task Mapping through Object-Oriented Task Description}},
booktitle = {Proceedings of the 1st Workshop on Transfer in Reinforcement Learning (TiRL) at AAMAS},
year = {2017}
}

This project was built on Python 2.7. Standard Python should be enough to execute the experiments, the graph generation code was originally published at https://github.com/f-leno/AdHoc_AAMAS-17 and for that you will need to install Jupyter Notebook (http://jupyter.readthedocs.io/en/latest/install.html).

The folder prey-predator contains all the Python source files.

The folders log and qtables contain pre-executed experiments that were used in the paper

The folder code stores all implementations. The expScript.py file executes all the experiments.

Executing again all experiments will take a long time. To print again all the graphs using the pre-generated result files, execute the file evaluation-leno.ipynb using jupyter notebook.

For questions about the Codification or paper, please send an email to the first author.