I work on information fusion and reasoning with uncertainty, mostly through the theory of belief functions.

I am a professor at Université d'Artois, France. I do my research in the LGI2A laboratory and my teaching at the IUT de Béthune, in the Networks & Telecommunications department. I have been a board member of the Belief Functions and Applications Society since 2012.

I graduated in 2003 with a DEA in artificial intelligence from Université Paul Sabatier, Toulouse, and earned a PhD from the Université de Technologie de Compiègne in 2006, carried out at Solystic, a world leader in mail-sorting solutions.

2019 –
Professor, Université d'Artois
2015
Habilitation (HDR), Université d'Artois
2008 – 2019
Associate professor, Université d'Artois
2006 – 2008
Assistant professor (ATER), UTC then Université d'Artois
2006
PhD, UTC, in partnership with Solystic
2003
DEA in artificial intelligence, Université Paul Sabatier, Toulouse

Research

4 threads

Correcting belief functions

Contextual discounting and reinforcement, canonical decompositions, and learning correction parameters from labelled or partially labelled data.

Evidential neural networks

Extending evidential classifiers (ENN, RBFN) with contextual corrections and dependence between positive and negative evidence.

Fault diagnosis of electrical machines

Fusing external flux sensors and random-convolution-kernel classifiers to detect stator faults, with the LSEE laboratory.

Uncertain data in applications

Vehicle routing with evidential demands, information exchange in vehicular networks, multi-target tracking, and stochastic simulation policies.

Publications

Theses

  • 2015
    Habilitation (HDR)
    Corrections et fusion d'informations dans le cadre des fonctions de croyance. Applications.
    Defended December 4, 2015 · Supervisor: É. Lefèvre
    pdfslides
  • 2006
    PhD thesis
    Fusion d'informations pour la reconnaissance automatique d'adresses postales dans le cadre de la théorie des fonctions de croyance.
    Defended December 6, 2006 · Supervisors: T. Denœux and M.-H. Masson
    pdfslidespres.

PhD students

2 in progress · 5 defended
  • 2023 –in progress
    Serigne Mamadou Diène
    Développement de réseaux de neurones profonds avec la théorie de Dempster-Shafer.
    Supervisor: D. Mercier (30%) · Co-supervisors: F. Pichon (35%), S. Ramel (35%) · Defense expected late 2026
  • 2023 –in progress
    Mouhamadou Mansour Lo
    Diagnosing motor/converter faults in an electric powertrain using machine learning and information fusion algorithms.
    Supervisor: F. Morganti (25%) · Co-supervisors: D. Mercier (25%), G. Morvan (25%), M. Rossi (25%) · Defense expected late 2026
  • 2021
    Siti Mutmainah
    Learning to adjust an evidential source of information using partially labeled data and partial decisions.
    Supervisor: D. Mercier (30%) · Co-supervisors: S. Hachour (35%), F. Pichon (35%) · Defended September 7, 2021
  • 2021
    Yu-Lin Huang
    Une nouvelle politique d'exécution de simulations stochastiques fondée sur des principes de partitionnement, de sélection et de clonage.
    Supervisor: D. Mercier (30%) · Co-supervisors: G. Morvan (35%), F. Pichon (35%) · Defended July 6, 2021
    pdfslidesdemo
  • 2017
    Pauline Minary
    Evidential calibration and fusion of multiple classifiers: application to face blurring. Industrial partnership with SNCF.
    Supervisor: É. Lefèvre (30%) · Co-supervisors: F. Pichon (35%), D. Mercier (35%) · Defended December 8, 2017
    pdfslidesdemo (video)
  • 2015
    Samir Hachour
    Suivi et classification d'objets multiples : contributions avec la théorie des fonctions de croyance.
    Supervisor: F. Delmotte (60%) · Co-supervisor: D. Mercier (40%) · Defended June 5, 2015
    pdfslides
  • 2014
    Mira Bou Farah
    Méthodes utilisant des fonctions de croyance pour la gestion des informations imparfaites dans les réseaux de véhicules.
    Supervisor: F. Delmotte (30%) · Co-supervisor: D. Mercier (70%) · Defended December 2, 2014
    pdfslides

Talks

selection
  • 2022
    Tutorials: introduction to belief functions and to their applications in classification
    With É. Lefèvre · Plate-Forme Intelligence Artificielle (PFIA), Saint-Étienne, June 27 – July 1, 2022
    intro BFintro classification
  • 2017
    On belief function corrections
    4th School on Belief Functions and Their Applications, Xi'an, China, July 5–9, 2017
    pdf
  • 2010
    Autour des mécanismes de correction de fonctions de croyance
    Réunion GDR ISIS — Avancées en fusion de données, February 11, 2010
    pdf
  • 2009
    Fusion d'informations imparfaites
    Petit Déjeuner Scientifique, IRSAA, March 31, 2009 — general-audience talk
    pdf
  • 2007
    Le Modèle des Croyances Transférables : une interprétation de la théorie des fonctions de croyance
    Séminaire LGI2A, Béthune, November 13, 2007
    pdf

Teaching

IUT de Béthune · R&T

I teach networks, web development and programming in the Networks & Telecommunications department (BUT R&T) of the IUT de Béthune. Course material is on Moodle Artois for enrolled students.

BUT R&TProgramming fundamentals
BUT R&TIntroduction to web development
BUT R&TInternet technologies, TCP/IP networks
BUT R&TNetwork security

Contact

E-mail and phone are on my laboratory page.

IUT de BéthuneDépartement R&T
1230 rue de l'Université, CS 20819
62408 Béthune Cedex, France
LGI2ALaboratoire de Génie Informatique et d'Automatique de l'Artois
Faculté des Sciences Appliquées, Technoparc Futura
62400 Béthune, France
© David Mercier · Terms of use Last update September 2026