Skip to content

DES-Lab/Automata-Learning-meets-Shielding

Repository files navigation

Automata Learning Meets Shielding

This repository contains source code and experiment data for paper "Automata Learning Meets Shielding".

Install

Clone and install python dependencies.

git clone git@github.com:DES-Lab/Automata-Learning-meets-Shielding.git
pip install -r requirements.txt

Install TEMPEST

# To install STORM
sudo apt-get install build-essential git cmake libboost-all-dev libcln-dev libgmp-dev libginac-dev automake libglpk-dev libhwloc-dev libz3-dev libxerces-c-dev libeigen3-dev
# To install TEMPEST
git clone https://github.com/tempest-shields/tempest-shields
cd tempest-shields
mkdir build
cd build
cmake ..
# building the command line interface for tempest suffices:
make storm-main
# if you have at least 8GB of RAM and multiple cores available you can speed up the build step via
make storm-main -j${NUMBER_OF_CORES_TO_BE_USED}

Run Experiments

Most of the code is based in the q_learning.py file. There you can find implementation of q-learning and shielded q-learning.

To reproduce experiments from the paper run:

./wrapper_q_learning.sh

About

Shielding RL agent based on model learned by automata learning

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published