AI Conference Deadlines: Stress Warning!

^&#8270 - equal contributions

Conference Proceedings

• UCB momentum Q-learning: Correcting the bias without forgetting.

  Pierre Ménard, Omar Darwiche Domingues, Xuedong Shang, and Michal Valko. In Proceedings of the 38th International Conference on Machine Learning, Vienna, Austria (ICML 2021). [long presentation - 3.2% acceptance rate]

  arxiv talk video
• Gamification of pure exploration for linear bandits.

  Rémy Degenne, Pierre Ménard, Xuedong Shang, and Michal Valko. In Proceedings of the 37th International Conference on Machine Learning, Vienna, Austria (ICML 2020).

  arxiv talk video
• Fixed-confidence guarantees for Bayesian best-arm identification.

  Xuedong Shang, Rianne de Heide, Emilie Kaufmann, Pierre Ménard and Michal Valko. In Proceedings of the 23rd International Conference on Artificial Intelligence and Statistics, Palermo, Italy (AIStats 2020).

  arxiv talk video
• General parallel optimization without a metric.

  Xuedong Shang, Emilie Kaufmann, and Michal Valko, In Proceedings of the 30th International Conference on Algorithmic Learning Theory, Chicago, USA (ALT 2019).

  hal talk

Journals

TBA

Workshop Papers

• Stochastic bandits with vector losses: Minimizing \(\ell^\infty\)-norm of relative losses.

  Xuedong Shang, Han Shao, and Jian Qian. In 15th European Workshop on Reinforcement Learning, Milan, Italy (EWRL 2022).

  arxiv poster
• A simple dynamic bandit algorithm for hyper-parameter tuning.

  Xuedong Shang, Emilie Kaufmann, and Michal Valko. In 6th ICML Workshop on Automated Machine Learning, Long Beach, USA (ICML 2019 - AutoML).

  hal poster code
• Adaptive black-box optimization got easier: HCT only needs local smoothness.

  Xuedong Shang, Emilie Kaufmann, and Michal Valko. In 14th European Workshop on Reinforcement Learning, Lille, France (EWRL 2018).

  hal poster

Software

• rlberry - A reinforcement learning library for research and education.

  Omar Darwiche Domingues^&#8270, Yannis Flet-Berliac^&#8270, Edouard Leurent^&#8270, Pierre Ménard^&#8270, Xuedong Shang^&#8270, and Michal Valko, GitHub repository (2021).

  demo code
• LinBAI.jl.

  Xuedong Shang, GitHub repository (2021).

  code

Preprints/Technical Reports/Working Papers

• Xuedong Shang, Emilie Kaufmann, and Michal Valko, Simple (dynamic) bandit algorithms for hyper-parameter optimization. Preprint (2019).
• Dominique Barbe, Alexandre Debant, Xuedong Shang, Time series clustering. Technical Report (2016).

Thesis

• PhD Thesis: 10/2017-03/2021, Team SequeL (SCOOL), Inria Lille-Nord Europe, Lille, France, Méthodes Adaptatives pour l'Optimisation dans un Environnement Stochastique, under the supervison of Emilie Kaufmann & Michal Valko.
• Internship & Master Thesis: 02/2017-07/2017, Team SequeL, Inria Lille-Nord Europe, Lille, France, Hierarchical Bandits for Black-Box Optimization and Monte-Carlo Tree Search, under the supervison of Emilie Kaufmann & Michal Valko.
• Internship: 05/2016-07/2016, Yamamoto-Cuturi Lab., Graduate School of Informatics, Kyoto University, Kyoto, Japan, Optimal transport geometry for sentiment analysis, under the supervision of Marco Cuturi.
• Internship: 06/2015-07/2015, Team Magnet, Inria Lille-Nord Europe, Lille, France, Recommender system (in French), under the supervision of Marc Tommasi.

Talks

• PhD Defense, Lille, France, September 29, 2021.
• Talk at RIKEN AIP, Tokyo, Japan, December 17, 2020.
• Talk at Seminar of Inria SequeL Team, Lille, France, February 29, 2020.
• Talk at Huawei Noah Ark's Lab, Paris, France, October 17, 2019.
• Talk at PhD Students' Seminar of Inria SequeL Team, Lille, France, February 2, 2017.

Services

• External Reviewer - NIPS 2017; ICML 2018, 2019, 2020; AIStats 2019
• Reviewer for Conferences - NeurIPS 2019^*, 2020^**, 2021, 2022; ICML 2021^†, 2023; ICLR 2021, 2023; AutoML-Conf 2022
• Reviewer for Workshops - ICLR 2019 LLD Workshop; ICML 2021 AutoML Workshop
• Reviewer for Journals - Mathematics of Operations Research, Econometrica
• PC Member - ICML 2022**
• Organization Committee - RLSS 2019
• Volunteer - ICLR 2020, ICML 2020, NeurIPS 2020

* - top reviewer, ** - best/outstanding reviewer, † - expert reviewer