Current Areas

  • Machine Learning for Mixed Integer Linear Programming (MILP):
    • Design of strategies using genetic programming to generate human-readable and generalizable algorithms.
    • Structural analysis of MILP instances using mathematical distance measures and clustering.
  • Sensor Networks & IoT:
    • Similarity-based scheduling and energy-efficient monitoring in large-scale, constrained sensor networks.

Publications

International Journals

  • P. Maillé, G. Maudet, M. Simon, B. Tuffin, “Are Search Engines Biased? Detecting and Reducing Bias using Meta Search Engines,” Electronic Commerce Research and Applications, 2022. (SJR: Q1) DOI

International Conferences

  • G. Maudet, G. Danoy,”A Distance Metric for Mixed Integer Programming Instances”, European Conference on Artificial Intelligence, 2025. (CORE: A) preprint
  • G. Maudet, G. Danoy, “Search Strategy Generation for Branch and Bound Using Genetic Programming,” Association for the Advancement of Artificial Intelligence, 2025. (CORE: A*) Link
  • G. Maudet, M. Batton-Hubert, P. Maillé, L. Toutain, “Energy Efficient Message Scheduling with Redundancy Control for Massive IoT Monitoring,” IEEE WCNC, 2023. (CORE: B) IEEE Xplore
  • G. Maudet, M. Batton-Hubert, P. Maillé, L. Toutain, “Emission Scheduling Strategies for Massive-IoT: Implementation and Performance Optimization,” IEEE/IFIP NOMS, 2022. (CORE: B) IEEE Xplore

National Conferences

  • G. Maudet, M. Batton-Hubert, P. Maillé, L. Toutain, “Grouper les Capteurs Similaires Grâce à leurs Données dans le Contexte de Massive IoT,” ALGOTEL, May 2024.
    HAL
  • G. Maudet, M. Batton-Hubert, P. Maillé, L. Toutain, “Réduction de la Redondance de Messages des Capteurs dans un Contexte Massive IoT,” ALGOTEL, May 2023.
    HAL

Ongoing Work

  • “Grouping Sensors Based on Observations in a Massive IoT Deployment,” preprint, HAL.
  • “A Survey On Data Collection Based on Sensors Similarity,” In progress.

PhD Thesis

Research Projects

SMILP: Structuring the Mixed Integer Linear Programming Space (Under review)
Lead author, University of Luxembourg. PI: Dr. Danoy.
Submitted to the FNR “CORE” project call, to fund a 2-year postdoc and a 4-year PhD.

SMILP aims to improve ML-based MILP solvers by introducing structural similarity metrics and clustering, enabling portfolio-based solving strategies with specialized ML models for instance groups. The project will develop hybrid learning methods and integrate with open-source tools (SCIP solver, MIPLIB instance library).