My name is Firas Safadi. I obtained a PhD in computer science from the University of Liège in 2015. I like artificial intelligence and video games.


I worked closely with Damien Ernst and Raphael Fonteneau. My research focused on the design of artificial intelligence for video games. I also have experience with various machine learning algorithms.

Parallel Cascade Correlation

I made a parallel implementation of the cascade correlation algorithm for GPU in CUDA C++. You can grab the code here. You can also view a short discussion here.

MASH Project

During my visit at the University of Paris-Sud 11 in Orsay in 2011, I worked on the MASH project in collaboration with Olivier Teytaud, Jean-Baptiste Hoock and Nataliya Sokolovska. My main contribution to the project is the implementation of a fitted Q iteration planner.


PhD Thesis Presentation
University of Liège
May 2015
Bitonic Sort
Circuit de Spa-Francorchamps
October 2014
Fitted Q Iteration
University of Paris-Sud 11
June 2011
Master's Thesis Presentation
University of Liège
June 2010


Artificial Intelligence in Video Games: Towards a Unified Framework
F. Safadi, R. Fonteneau, D. Ernst
International Journal of Computer Games Technology
March 13th, 2015
Imitative learning for real-time strategy games
Q. Gemine, F. Safadi, R. Fonteneau, D. Ernst
8th IEEE Conference on Computational Intelligence and Games (CIG 2012)
Granada, Spain
September 11-14, 2012
Artificial intelligence design for real-time strategy games (Poster)
F. Safadi, R. Fonteneau, D. Ernst
Decision Making with Multiple Imperfect Decision Makers
25th Annual Conference on Neural Information Processing Systems (NIPS 2011)
Sierra Nevada, Spain
December 16th, 2011