Research

Research Interests

  • Fitness Landscape Analysis
  • Adaptive approaches (local search algorithms, heuristics...)
  • Automatic configuration of algorithms
  • Multi-objective Optimization

International collaborations

  • Meta-learner for fast parameter-free algorithms

Myriam Delgado -- University of Technology of Paraná (Brazil)

  • MO-ParamILS: A Multi-objective Automatic Algorithm Configuration Framework

Holger Hoos -- University of British Columbia (Vancouver, Canada)
Heike Trautmann -- University of Münster (Germany)

  • General framework to configure single-objective local search algorithms automatically

Franco Mascia and Thomas Stützle -- Université Libre de Bruxelles (IRIDIA team, Belgium)
Manuel López-Ibánez -- University of Manchester (UK)

  • Neutrality in multi-objective optimization: definitions and interpretations

Hernán Aguirre and Kyoshi Tanaka -- Shinshu University (Nagano, Japan)

  • Factoradics representation for permutation problems in local search algorithms

Olivier Regnier-Coudert -- Gordon University (Aberdeen, UK)

  • MO-ParamILS: A Multi-objective Automatic Algorithm Configuration Framework

Holger Hoos -- University of Leiden (The Netherlands)

Supervision

PhD thesis

  • Laurent Parmentier (2019-...)
    Multi-objective automatic design of machine learning systems
    HDR : Laetitia Jourdan
  • Camille Pageau (2018-...)
    Optimization for automatic design and configuration of datamining algorithms: Application to medical data
    HDR : Laetitia Jourdan
  • Weerapan Sae-dan (2018-...)
    Automatic Design of Dynamic Local Search Algorithms
    HDR : Laetitia Jourdan
  • Lucas Marcondes Pavelski (2017-...)
    Meta-learning and transfer-learning for the flow shop problem
    with Myriam Delgado (University of Technology of Paraná , Brazil)
  • Lucien Mousin (2015-2018)
    Exploiting knowledge to improve performance of optimization approaches
    HDR : Clarisse Dhaenens
  • Aymeric Blot (2015-2018)
    Designing adaptive algorithms for multi-objective combinatorial optimization problems
    HDR : Laetitia Jourdan

Industrial collaborations

  • PIXEO : supervision of Anne-Lise Bedenel (engineer), May 2014-May 2015
    Data analysis and predictive model

Project

  • Interreg V France-Wallonie-Vlaanderen - PATHACOV – 2018-2022
  • PHRCI-16-089 CATOCOV
  • Projet ANR ClinMine: Optimisation de la prise en Charge des Patients à l'Hôpital

Internships

Master thesis

  • Camille Pageau (Stage de fin d’étude, Polytech Lille, Université de Lille)
    Automatic algorithm configuration applied to planification of strategies, March-August 2018 (6 months)
  • Lucien Mousin (Master 2, Université Lille 1)
    Designing a multi-objective local search with a trail mechanism for features selection, March-August 2015 (6 months)
  • Aymeric Blot (Master 2, ENS Cachan/Bretagne, Université de Rennes)
    Design of Multi-objective Local Search Algorithms using Structure Features. Application to the Permutation Flowshop Scheduling Problem, February-July 2014 (6 months)

Undergraduate

  • Clément Castelein (5th year engineer -- Polytech Lille, Université de Lille)
    Automatic Configuration of the Clarke and Wright algorithm, October 2018-February 2019
  • Camille Pageau (5th year engineer -- Polytech Lille, Université de Lille)
    Automatic algorithm configuration applied to planification of strategies, October 2017-February 2018
  • Ismahane Ben Gayed (5th year engineer -- Polytech Lille, Université de Lille)
    Analysis of paramaters of the Clarke and Wright algorithm, October-December 2017
  • Matthieu Bernard (3rd year engineer -- Polytech Lille, Université de Lille)
    Landscape Analysis of the no-wait flowshop scheduling problem, June 2017
  • Mariette Biernacki (3rd year engineer -- Polytech Lille, Université de Lille)
    'Extracting knowledge of permutation flowshop
  • Christopher Lievin (Stage IUT)

Development of a software to present the no-wiat flowshop, April-June 2016

  • Thomas Dubrulle (5th year engineer, Polytech'Lille, Université Lille 1)
    Multi-objective Knapsack Problem: Fitness Landscape Analysis and performance analysis of stochastic approaches, project = 100h, October 2015-February 2016
  • Thomas Dubrulle (4th year engineer, Polytech'Lille, Université Lille 1)
    Multi-objective Local search for TSP, May-July 2015 (3 months)
  • Adam Pamart (5th year engineer, Polytech'Lille, Université Lille 1)
    Comparison of automatic tuning tools for optimization algorithms, project = 100h, January-February 2015
  • Lucien Mousin (Master 2, Université Lille 1)
    Adapting a tabu search with a trail mechanism for features selection, project = 100h, October 2014-January 2015
  • Marion Peycelon and Marion Lambert (5th year engineer, Polytech'Lille, Université Lille 1)
    Adapting a tabu search with a trail mechanism for scheduling, project = 100h, October 2013-February 2014
  • Aymeric Blot (Bachelor 3, ENS Cachan/Bretagne, Université de Rennes)
    Analysis and exploitation of the neutrality of the graph coloring problem, June-July 2012 (6 weeks)
  • Pierre Dezoomer (5th year engineer, Polytech'Lille, Université Lille 1)
    Analysis of the influence of benchmarks generation on the difficulty of created optimization problems, March-April 2011 (6 weeks)

PhD thesis (2008 - 2011)

[details]

The thesis and the slides of the defense are available here

      

      

      

Blix powered by PmWiki