2023

• Toward efficient then explainable models: Choquetization of a Neural Net.
  • Details: St_Explain.pdf
• Stability-based causal discovery: an adversarial approach
  • Details:St_Stable.pdf

Below is the list of internships until 2022.
Please check the new "Job Offers" page of the TAU team

Propositions de stages Master Recherche ou Ecole d'Ingénieurs - Année scolaire 2021-2022

• The Game of Privacy CLOSED
  • Contact: Michele Sebag, Armand Lacombe at lri dot fr
  • Topic: Supervised Machine Learning, Reinforcement Learning, Privacy
  • Details: Sujet_Privacy.pdf

• Combatting Unemployment: Data, Recommendations and Fairness
  • Contact: Michele Sebag, Philippe Caillou, sebag, caillou at lri dot fr
  • Topic: Machine Learning, Optimal Transport, Recommendations, Causal Learning
  • Details:see here.

• Spatio-temporal causal learning: Health outcomes of residential pesticides
  • Contact: Michele Sebag at lri dot fr; Olivier Allais at inrae dot fr
  • Topic: Machine Learning, Causal Learning, Generative models
  • Details:Sujet_Horapest.pdf

• Learning to Run a Power Network Energies of the future and carbon neutrality CLOSED
  • Contact: Isabelle Guyon, iguyon at lri dot fr
  • Topic: Machine Learning, Reinforcement learning, Power consumption, Benchmark
  • Details:see here.

• Meta-Album: creation of a meta-learning benchmark CLOSED
  • Contact: Isabelle Guyon, iguyon at lri dot fr
  • Topic: Machine Learning, Cross-Domain Meta-learning, Computer Vision, Benchmark
  • Details:see here.

• Robustness against bias in AutoDL systems CLOSED
  • Contact: Isabelle Guyon, iguyon at lri dot fr
  • Topic: Machine Learning, bias in data, deep learning
  • Details:see here.

Propositions de stages Master Recherche ou Ecole d'Ingénieurs - Année scolaire 2020-2021

• Recherche de politique optimale dans un espace 1D ou 2D, Application à l’agriculture durable
  • Contact: Simon Moulieras, simon.moulieras at greenshield.fr, Michele Sebag, sebag at lri dot fr
  • Topic: Reinforcement Learning, Graph Neural Networks
  • Details:ici.

• Post-hoc Model-Agnostic Explanation through Evolutionary Computing-based Methods
  • Contact: Marc Schoenauer, Marc dot Schoenauer at inria dot fr
  • Topic: Explainable and Trustworthy Machine Learning, Evolutionary Computation
  • Details:see here.

• Is overfitting avoidance “all you need” to guarantee privacy? CLOSED
  • Contact: Isabelle Guyon, iguyon at lri dot fr
  • Topic: Protection of privacy, synthetic data, GANs
  • Details:see here.

• REVEAL games, a new class of RL problems CLOSED
  • Contact: Isabelle Guyon, iguyon at lri dot fr
  • Topic: Reinforcement learning, meta-learning
  • Details:see here.

• Creation of a learning model for photoaerial identification of potential old-growth forests CLOSED
  • Contact: Isabelle Guyon, iguyon at lri dot fr
  • Topic: Computer vision, ecology
  • Details:see here.

• Controlling the COVID epidemic with reinforcement learning CLOSED
  • Contact: Isabelle Guyon, iguyon at lri dot fr
  • Topic: Reinforcement learning, epidemiology
  • Details:see here.

• (Deep) Graph Neural Networks for Fundamental Physics: the case of Glassy liquids
  • remark: This position has been filled ! (others may have been as well, this one it's for sure
  • Contact: François Landes, francois dot landes at inria dot fr
  • Topic: Graph Neural Networks (GNN), simulation data, fundamental physics
  • Details:see here.

Propositions de stages Master Recherche ou Ecole d'Ingénieurs - Année scolaire 2019-2020

• Exploiting heterogeneous COVID-databases with limited manual effort: a domain adaptation approach
  • Contact : Michele Sebag, sebag at lri dot fr (UPSaclay/CNRS/INRIA)
    avec Francois Landes (UPSaclay & Inria) and Isabelle Guyon (UPSaclay&Inria).
  • Sujet_Wrang.pdf
    Topic: Machine Learning, Data Wrangling, Domain Adaptation, Data cleaning

• Automated Deep Learning Self-Service Against COVID-19 CLOSED
  • Contact : Isabelle Guyon, iguyon at lri dot fr
    avec Zhengying Liu (doctorant), Adrien Pavao (doctorant) et en collaboration avec Google Zurich
  • Details
    La competition AutoDL a produit ses fruits et livré des solutions qui peuvent automatiquement entrainer des classifieurs puissants pouvant résoudre des problèmes de computer vision, text processing, et speech *sans intervention humaine aucune*. Vous les mettrez à disposition du public pour résoudre des problèmes liés au COVID-19.

• COVID-19 Facts and Rumors CLOSED
  • Contact : Isabelle Guyon, iguyon at lri dot fr (UPSaclay/INRIA)
    avec Ioana Manolescu-Goujot (INRIA Saclay and LIX/Ecole Polytechnique), Zhengying Liu (UPSaclay), Adrien Pavao (UPSaclay).
  • https://docs.google.com/document/d/1j0FEh5PR3bNP_qtsa-HugdWfH89zmJRnQpyFskhrkG0/edit?usp=sharing
    Topic: Machine Learning, Fact Checking, Statistics
    Stage de master uniquement.

• COVID-19 risk mitigation CLOSED
  • Contact : Isabelle Guyon, iguyon at lri dot fr (UPSaclay/INRIA)
    avec Michele Sebag (UPSaclay/CNRS), Kristin Bennett (RPI), John Erickson (RPI), Paloa Tubaro (UPSaclay/CNRS), Adrien Pavao (UPSaclay).
  • https://docs.google.com/document/d/1RNgucM3AgXcY4OI4-0BEHiSnmr-iqK5dUJvTCN2vzQ0/edit?usp=sharing
    Topic: Machine Learning, Causality, Epidemiology, Econometrics, Optimization
    Stage de master uniquement.

• COVID-19, coping with confinement CLOSED
  • Contact : Isabelle Guyon, iguyon at lri dot fr (UPSaclay/INRIA)
    avec Nicolas Thiery (UPSaclay), Sergio Escalera (U. Barcelona), Alice Guyon (U. Côte d’Azur), Adrien Pavao (UPSaclay).
  • [https:/%2