laurent

Laurent Charlin, Ph.D.

Assistant professor at HEC Montréal
(also adjunct in CS by courtesy).

Canada CIFAR AI Chair

Email: lcharlin [at] gmail [dot] com


 

[bio]  [research group]  [papers]  [projects]  [courses]  [talks]  [press]  [CV]

 

News:


My Interests:

I develop novel machine learning models, particularly probabilistic graphical models, to help in decision making. My recent work has focussed on extending the capabilities of recommender and dialog systems.

I am generally interested in applying learning methods to analyze different data.


Short Bio:

[Current] Assistant Professor of statistics at HEC Montréal (U. Montreal's business school).

Postdoc at McGill University. Advised by Joelle Pineau.

Postdoc at Princeton University and Columbia University. Advised by David Blei.

Ph.D. in the machine learning group at the University of Toronto. Advised by Richard Zemel and Craig Boutilier.

Master's at the University of Waterloo. Advised by Pascal Poupart.


Research Group:

Alumni:


Papers:

    Pre prints

  • Inference for travel time on transportation networks
    Mohamad Elmasri, Aurelie Labbe, Denis Larocque, Laurent Charlin
    [arXiv]

  • IG-RL: Inductive Graph Reinforcement Learning for Massive-Scale Traffic Signal Control
    François-Xavier Devailly, Denis Larocque, Laurent Charlin
    To appear in the IEEE Transactions on Intelligent Transportation Systems
    [arXiv]

    Publications

  • Online Fast Adaptation and Knowledge Accumulation: a New Approach to Continual Learning
    Massimo Caccia, Pau Rodriguez, Oleksiy Ostapenko, Fabrice Normandin, Min Lin, Lucas Caccia, Issam Laradji, Irina Rish, Alexande Lacoste, David Vazquez, Laurent Charlin
    In the proceedings of NeurIPS 2020
    [arXiv]

  • Synbols: Probing Learning Algorithms with Synthetic Datasets
    Alexandre Lacoste, Pau Rodríguez, Frédéric Branchaud-Charron, Parmida Atighehchian, Massimo Caccia, Issam Laradji, Alexandre Drouin, Matt Craddock, Laurent Charlin, David Vázquez
    In the proceedings of NeurIPS 2020
    [arXiv]

  • Multi-XScience: A Large-scale Dataset for Extreme Multi-document Summarization of Scientific Articles
    Yao Lu, Yue Dong, Laurent Charlin
    In the proceedings of EMNLP 2020 (Short Paper)
    [paper] [Dataset]

  • On the effectiveness of two-step learning for latent-variable models
    Cem Subakan, Maxime Gasse, Laurent Charlin
    In the proceedings of MLSP 2020
    [paper]

  • Causal Inference for Recommender Systems
    Yixin Wang, Dawen Liang, Laurent Charlin, David M. Blei
    In the proceedings of ACM RecSys 2020 (Short Paper)
    [paper] [arXiv (longer version)]

  • Language GANs Falling Short
    Massimo Caccia, Lucas Caccia, William Fedus, Hugo Larochelle, Joëlle Pineau, Laurent Charlin
    In the proceedings of ICLR 2020
    [arXiv]

  • Exact Combinatorial Optimization with Graph Convolutional Neural Networks
    Maxime Gasse, Didier Chételat, Nicola Ferroni, Laurent Charlin, Andrea Lodi
    In the proceedings of NeurIPS 2019
    [arXiv] [code]

  • Online Continual Learning with Maximal Interfered Retrieval
    Rahaf Aljundi, Lucas Caccia, Eugene Belilovsky, Massimo Caccia, Min Lin, Laurent Charlin, Tinne Tuytelaars
    In the proceedings of NeurIPS 2019
    [arXiv]

  • Continual Learning of New Sound Classes using Generative Replay
    Zhepei Wang, Cem Subakan, Efthymios Tzinis, Paris Smaragdis, Laurent Charlin
    In the Proceedings of WASPAA 2019
    [arXiv]

  • Session-based Social Recommendation via Dynamic Graph Attention Networks
    Weiping Song, Zhiping Xiao, Yifan Wang, Laurent Charlin, Ming Zhang and Jian Tang
    In the proceedings of WSDM 2019.
    [paper]

  • Towards Deep Conversational Recommendations
    Raymond Li, Samira Ebrahimi Kahou, Hannes Schulz, Vincent Michalski, Laurent Charlin, Christopher Pal
    In the proceedings of NeurIPS 2018
    [paper] [project website (including dataset)]

  • Focused Hierarchical RNNs for Conditional Sequence Processing
    Nan Rosemary Ke, Konrad Żołna, Alessandro Sordoni, Zhouhan Lin, Adam Trischler, Yoshua Bengio, Joelle Pineau, Laurent Charlin, Chris Pal
    In the proceedings of ICML 2018
    [paper]

  • A Survey of Available Corpora For Building Data-Driven Dialogue Systems: The Journal Version
    Iulian Vlad Serban, Ryan Lowe, Peter Henderson, Laurent Charlin, Joelle Pineau.
    (Journal) Dialogue and Discourse, 2018
    [paper]

  • Training End-to-End Dialogue Systems with the Ubuntu Dialogue Corpus
    Ryan Lowe, Nissan Pow, Iulian Vlad Serban, Laurent Charlin, Chia-Wei Liu, Joelle Pineau
    (Journal) Dialogue and Discourse, 2017
    [paper]

  • A Hierarchical Latent Variable Encoder-Decoder Model for Generating Dialogues
    Iulian Vlad Serban, Alessandro Sordoni, Ryan Lowe, Laurent Charlin, Joelle Pineau, Aaron Courville, Yoshua Bengio
    In the proceedings of AAAI 2017
    [paper] [code]

  • How NOT To Evaluate Your Dialogue System: An Empirical Study of Unsupervised Evaluation Metrics for Dialogue Response Generation
    Chia-Wei Liu, Ryan Lowe, Iulian Vlad Serban, Michael Noseworthy, Laurent Charlin, Joelle Pineau.
    In the proceedings of EMNLP 2016
    [arXiv]

  • Factorization Meets the Item Embedding: Matrix Factorization with Item Co-occurrence
    Dawen Liang, Jaan Altosaar, Laurent Charlin, David M. Blei.
    In the proceedings of Recsys 2016
    [pdf] [code]

  • Modeling User Exposure in Recommendation
    Dawen Liang, Laurent Charlin, James McInerney, David M. Blei.
    In the proceedings of WWW 2016
    [paper] [code]

  • Dynamic Poisson Factorization
    Laurent Charlin, Rajesh Ranganath, James McInerney, David M. Blei.
    In the proceedings of RecSys 2015
    [paper] [publication source] [code] [slides]

  • Deep Exponential Families
    Rajesh Ranganath, Linpeng Tang, Laurent Charlin, David M. Blei.
    In the proceedings of AISTATS 2015
    [paper] [appendix] [code] [publication source]

  • Content-based recommendations with Poisson factorization
    Prem Gopalan, Laurent Charlin, David M. Blei.
    In the proceedings of NIPS 2014.
    [paper] [appendix] [code] [publication source]

  • Leveraging User Libraries to Bootstrap Collaborative Filtering
    Laurent Charlin, Richard Zemel, Hugo Larochelle.
    In the proceedings of KDD 2014.
    [paper][video lecture]

  • The Toronto Paper Matching System: An automated paper-reviewer assignment system
    Laurent Charlin, Richard Zemel.
    In the ICML Workshop on Peer Reviewing and Publishing Models (PEER), 2013.
    [paper]

  • Stochastic k-Neighborhood Selection for Supervised and Unsupervised Learning
    Daniel Tarlow, Kevin Swersky, Laurent Charlin, Ilya Sutskever, Richard Zemel.
    In proceeding of the International Conference on Machine Learning (ICML), 2013.
    [paper] [code]

  • Active Learning for Matching Problems
    Laurent Charlin, Richard Zemel, Craig Boutilier
    In proceedings of the International Conference on Machine Learning (ICML), 2012.

    [paper] [bibtex] [poster] [short video lecture]

  • A Framework for Optimizing Paper Matching
    Laurent Charlin, Richard Zemel, Craig Boutilier
    In proceedings of Uncertainty in Artificial Intelligence (UAI), Barcelona, 2011.

    [paper] [bibtex] [poster]

  • Hierarchical POMDP Controller Optimization by Likelihood Maximization -- Best paper award runner-up
    Marc Toussaint, Laurent Charlin, Pascal Poupart
    In proceedings of Uncertainty in Artificial Intelligence (UAI), Helsinki, 2008.
    [paper] [bibtex] [video lecture]

  • Hierarchical POMDP Controller Optimization by Likelihood Maximization
    Marc Toussaint, Laurent Charlin, Pascal Poupart
    In proceedings of the AAAI workshop on Advancements in POMDP Solvers, Chicago, 2008. (A longer version has also appeared in UAI'08, see above)
    [paper]

  • Automated Hierarchy Discovery for Planning in Partially Observable Environments
    Laurent Charlin, Pascal Poupart and Romy Shioda
    Advances in Neural Information Processing Systems 19 (NIPS), 2007.
    [paper ps] [paper ps.gz] [paper pdf] [bibtex]

  • MAXSM: A MultiHeuristic Approach to XML Schema Matching
    Mirza Beg, Laurent Charlin and Joel So
    University of Waterloo Technical Report, CS-2006-47, 2006.
    [paper pdf]
Master's Thesis:
  • Automated Hierarchy Discovery for Planning in Partially Observable Domains,
    Master's Thesis, School of Computer Science, University of Waterloo, December 2006.
    [ps] [ps.gz] [pdf] [bibtex]

Projects:

  • Rich Zemel and I have developed a system to help match reviewers to papers for conferences. The system is now used by several of the leading machine learning and computer vision conferences.

    For more information check out: Toronto Paper Matching System (TPMS)


Courses:


Press:

  • Our research about active learning for matching problems, and especially its links to online dating, was briefly featured in the UofT Magazine:

    Computer Says No


Selected Talks:

  • Graphical models for analyzing and understanding user behaviour, Ecole Polytechnique Montréal - 10/2016.
  • User Modelling, RecProfil workshop at RecSys'16, Boston - 09/2016.
  • Matrix factorization models for recommender systems, Twitter, Boston - 06/2016.
  • Recommender Systems: going beyond Matrix Factorization, Bloomberg, New York - 03/2015.
  • Recommender Systems: going beyond Matrix Factorization, McGill University - 02/2015.
  • Intelligent-environments group (CREI), Université de Sherbrooke - 02/2013.
  • AdaComp seminar, National University of Singapore (NUS) - Hierarchical POMDP Controller Optimization by Likelihood Maximization - 11/2008.
  • Reinforcement Learning seminar, McGill University - Hierarchical POMDP Controller Optimization by Likelihood Maximization - 11/2008.
  • Machine Learning Seminar, TU-Berlin, Germany - Automated Hierarchy Discovery for Planning in Partially Observable Environments - 07/2007.
  • MITACS Machine Learning Seminar, McGill University - Automated Hierarchy Discovery for Planning in Partially Observable Environments - 05/2007.
  • CoGS Seminars, University of Toronto, Automated Hierarchy Discovery for Planning in Partially Observable Environments - 02/2007.
  • AI seminar, University of Waterloo - Automated Hierarchy Discovery for Planning in Partially -01/2007.
Last updated: June 2019