Didier Chételat

I am a machine learning research scientist at the Canada Excellence Research Chair in Data Science for Real Time Decision Making, under the leadership of Andrea Lodi. I obtained my Ph.D. in Statistics at Cornell University in 2015 under the supervision of Martin T. Wells.
My main research interests center around deep learning for combinatorial optimization. I also work on applied projects, most recently in finance, transportation, medical imaging and physical engineering.

Software

Ecole    [code] [documentation] A Python library for Gym-like integration of machine learning in combinatorial optimization solvers.

Publications

• Continuous cutting plane algorithms in integer programming    [code]
  Operations Research Letters. To appear. Didier Chételat, Andrea Lodi.
• Combinatorial optimization and reasoning with graph neural networks (Extended version)
  Journal of Machine Learning Research 24, 1–61. Quentin Cappart, Didier Chételat, Elias B. Khalil, Andrea Lodi, Christopher Morris, Petar Veličković.

• Learning to branch with tree MDPs    [code]
  Advances in Neural Information Processing Systems 35, 18514–18526. Lara Scavuzzo, Feng Y. Chen, Didier Chételat, Maxime Gasse, Andrea Lodi, Neil Yorke-Smith, Karen Aardal.
• Learning to compare nodes in branch and bound with graph neural networks    [code]
  Advances in Neural Information Processing Systems 35, 32000–32010. Abdel G. Labassi, Didier Chételat, Andrea Lodi.
• Lookback for learning to branch
  Transactions on Machine Learning Research, November 2022. Prateek Gupta, Elias B. Khalil, Didier Chételat, Maxime Gasse, Yoshua Bengio, Andrea Lodi and P. M. Kumar.

• Combinatorial optimization and reasoning with graph neural networks
  International Joint Conference on Artificial Intelligence 30, 4348-4355. Quentin Cappart, Didier Chételat, Elias B. Khalil, Andrea Lodi, Christopher Morris, Petar Veličković.

• Ecole: A Gym-like library for machine learning in combinatorial optimization solvers    [code]
  Learning Meets Combinatorial Algorithms Neurips Workshop. Antoine Prouvost, Justin Dumouchelle, Maxime Gasse, Didier Chételat, Andrea Lodi.

• Exact combinatorial optimization with graph convolutional neural networks    [code]
  Advances in Neural Information Processing Systems 32, 15554–15566. Maxime Gasse, Didier Chételat, Nicola Ferroni, Laurent Charlin, Andrea Lodi.
• The middle-scale asymptotics of Wishart matrices
  Annals of Statistics 47, 2639-2670. Didier Chételat, Martin T. Wells.

• On the domain of attraction of a Tracy–Widom law with applications to testing multiple largest roots
  Journal of Multivariate Analysis 165, 132-142. Didier Chételat, Rajendran Narayanan, Martin T. Wells.

• Optimal two-step prediction in regression
  Electronic Journal of Statistics 11, 2519-2546. Didier Chételat, Joseph Salmon, Johannes Lederer.

• Improved second order estimation in the singular multivariate normal model
  Journal of Multivariate Analysis 147, 1-19. Didier Chételat, Martin T. Wells.

• Improved multivariate normal mean estimation with unknown covariance when p is greater than n
  Annals of Statistics 40, 3137-3160. Didier Chételat, Martin T. Wells.