Daniel Porumbel

Knowledge and Decision Team
Lina CS. Laboratory,
Nantes, France

Contact: daniel.porumbel AT univ - nantes DOT fr
Metaheuristics, Optimization and Applications Team
LERIA CS. Laboratory,
Angers, France
 
 
Just me

Refereed publications

Daniel Porumbel, Jin-Kao Hao and Pascale Kuntz, Reinforced Tabu Search for Graph Coloring. Submitted, 2007.

Daniel Porumbel, Jin-Kao Hao and Pascale Kuntz, A study of evaluation functions for the graph K-coloring problem. Presented at the 8th International Conference on Artificial Evolution and to appear in Lecture Notes in Computer Science, Springer, 2007. (ps) Springer© 

Daniel Porumbel, Jin-Kao Hao and Pascale Kuntz, Recherche Tabou renforcée pour la coloration de graphe. Accepted to the 9th congress of the French Operational Research Society, February 25-27, 2008. (ps, in french)


Other reseach and teaching materials

I maintain a Graph Coloring Page with the perfomances of the best algorithms on all instances from the second Dimacs Implementation Challenge.
Course list (an archive of teaching notes that I wrote up since 2004)
Other notes (interesting mathematical facts, computer science projects, etc.)

Quick Facts

I started my PhD in December 2006 under the supervision of two professors: Pascale Kuntz Cosperec in Nantes and Jin Kao Hao in Angers. I am interested in heuristic algorithms , combinatorial optimization and applying machine learning techniques to optimization. Apart from my main phd research subject, I have also carried out research on various other fields from pure mathematics (I am an International Mathematical Olympiad medalist) to DNA-based computers, multi-agent Systems, natural language processing computer graphics (annimated rendering) or Grid systems; see this archive.

  · M.S. Knowledge Extraction from Data, Ecole Polytechnique de l'Université de Nantes, a joint degree program with Université Lumiere Lyon 2, Université Paris Sud XI, France
  · B.S. University "Politehnica" of Bucharest, Romania, Computer Science and Engineering Department