Sylvain Arlot

Professor - Laboratoire de Mathématiques, Université Paris-Saclay, Orsay.

Département de Mathématiques
Bâtiment 307
Faculté des Sciences d'Orsay
Université Paris-Saclay
F-91405 Orsay Cedex
France
E-mail: FR
Office: Office 3A10 [3rd floor, IMO building --- Map --- Directions]
Phone: +33 1 69 15 67 31
Fax: +33 1 69 15 60 34
GPG public key (Key fingerprint 60A6 45C6 9CCA 10E4 B98C 4ECA 40C9 6F55 6AA2 D12F)

Teaching   -   Publications   -   Software   -   Talks   -   Short CV (English / French)   -   Miscellaneous



I am professor at the Mathematics Department of University Paris-Saclay (Faculté des Sciences d'Orsay), member of the Probability and Statistics team of the LMO. I am junior member of Institut Universitaire de France (IUF) for five years, starting September 2020.
I am head of the Celeste INRIA Saclay project-team.
I am associate editor for the Annales de l'Institut Henri Poincaré B, Probability and Statistics.
I co-organize the internal group meeting of Celeste, and I am member of the scientific committee of the Séminaire palaisien.


Main research interests -- Thèmes de recherche


Teaching -- Enseignement



Grants -- Financements


Publications & Preprints

Bibtex - Hal - Google Scholar
 
Preprints -- Prépublications Book chapters
  1. (2018) Sylvain Arlot. Fondamentaux de l'apprentissage statistique. In Myriam Maumy-Bertrand, Gilbert Saporta et Christine Thomas-Agnan, éditeurs, Journées d'Études en Statistique (JES) 2016: Apprentissage statistique et données massives, Éd. Technip. pages 17--139.
    [Éditeur]   [Hal]   [page dédiée]   [transparents]
  2. (2018) Sylvain Arlot. Validation croisée. In Myriam Maumy-Bertrand, Gilbert Saporta et Christine Thomas-Agnan, éditeurs, Journées d'Études en Statistique (JES) 2016: Apprentissage statistique et données massives, Éd. Technip. pages 141--174.
    [Éditeur]   [Hal]   [page dédiée]   [transparents]
Textbook (in French) -- Manuel d'exercices
  1. (2015) Sylvain Arlot, Aurélien Garivier and Gilles Stoltz. Exercices d'oral de mathématiques -- BL - ECE - ECS. Corrigés et commentés. Coll. Références sciences, Ellipses. 528 pages. Issu de notre participation au concours d'entrée à l'École normale supérieure, voie B/L.
Peer-reviewed published papers -- Articles publiés dans des journaux ou conférences à comité de lecture
  1. (2023) Pierre Humbert, Batiste Le Bars, Aurélien Bellet, Sylvain Arlot. One-Shot Federated Conformal Prediction. ICML 2023.
    [conference]   [pdf]   [arXiv]   [hal]   [code]   [reviews]   [poster+video]
  2. (2022) Binh T. Nguyen, Bertrand Thirion, Sylvain Arlot. A Conditional Randomization Test for Sparse Logistic Regression in High-Dimension. NeuRIPS 2022.
    [arXiv]   [hal]   [reviews]   [code]
  3. (2021) Guillaume Maillard, Sylvain Arlot, Matthieu Lerasle. Aggregated hold-out. Journal of Machine Learning Research 22(20):1--55, 2021.
    [journal]   [pdf]   [Hal]   [arXiv]   [slides Cambridge February 2019]
    See also the 2017 preprint (focused on classification): Cross-validation improved by aggregation: Agghoo. [Hal]   [arXiv]
  4. (2020) Binh T. Nguyen, Jérôme-Alexis Chevalier, Bertrand Thirion and Sylvain Arlot. Aggregation of Multiple Knockoffs. Proceedings of the 37th International Conference on Machine Learning (ICML 2020), PMLR 119.
    [ICML]   [arXiv]   [hal]   [python code]
  5. (2019) Sylvain Arlot, Alain Celisse, Zaïd Harchaoui. A kernel multiple change-point algorithm via model selection. Journal of Machine Learning Research 20 (162): 1--56, 2019.
    [journal]   [pdf]   [Hal]   [arXiv]   [python package, by Jones and Harchaoui]   [slides Lille april 2014]   [slides Nanterre november 2017]
  6. (2019) Sylvain Arlot. Minimal penalties and the slope heuristics: a survey. Journal de la Société Française de Statistique, Vol 160, No 3. 1--106.
    [journal]   [Hal]   [arXiv]   [slides CIRM december 2013]
    (2019) Sylvain Arlot. Rejoinder on Minimal penalties and the slope heuristics: a survey. Journal de la Société Française de Statistique, Vol 160, No 3. 158--168.
    [journal]   [Hal]   [arXiv]
    Full special issue, with discussion:   [journal]
  7. (2019) Sylvain Arlot, Stefano Marmi, Duccio Papini. Coupling the Yoccoz-Birkeland population model with price dynamics: chaotic livestock commodities market cycles. Nonlinearity 32 2564.
    [journal]   [Hal]   [arXiv]   [matlab code]
  8. (2018) Damien Garreau, Sylvain Arlot. Consistent change-point detection with kernels. Electronic Journal of Statistics 2018, Vol. 12, No. 2, 4440-4486.
    [journal]   [Hal]   [arXiv]   [slides Nanterre November 2017]
  9. (2016) Sylvain Arlot, Matthieu Lerasle. Choice of V for V-Fold Cross-Validation in Least-Squares Density Estimation. JMLR 17(208):1--50, 2016.
    [journal]   [pdf]   [appendix]   [Hal]   [arXiv]   [slides Nice february 2015]
  10. (2016) Sylvain Arlot, Robin Genuer. Comments on: A random forest guided tour. TEST. June 2016, Volume 25, Issue 2, pp 228-238.
    The final publication is available at Springer via this link.
    [journal]   [Hal]   [arXiv]   [slides October 2018]
    Related paper by Biau and Scornet:   [journal]   [arXiv]
  11. (2016) Rémi Lajugie, Piotr Bojanowski, Philippe Cuvillier, Sylvain Arlot and Francis Bach. A weakly-supervised discriminative model for audio-to-score alignment. 41st International Conference on Acoustics, Speech, and Signal Processing (ICASSP), March 2016, Shanghai, China.
    [Hal]   [Conference]
  12. (2014) Damien Garreau, Rémi Lajugie, Sylvain Arlot, Francis Bach. Metric Learning for Temporal Sequence Alignment. In Advances in Neural Information Processing Systems (NIPS) 27, 1817-1825.
    [proceedings]   [dedicated webpage in the proceedings]   [pdf]   [supplementary material]   [code on Github] [Hal]   [arXiv]
  13. (2014) Rémi Lajugie, Sylvain Arlot, Francis Bach. Large-Margin Metric Learning for Constrained Partitioning Problems. ICML 2014. JMLR W&CP 32 (1): 297-305.
    [proceedings]   [long version, Hal]   [long version, arXiv]
    See also this paper at the NIPS 2012 Workshop on Discrete Optimization in Machine Learning (DISCML) 2012.
  14. (2012) Matthieu Solnon, Sylvain Arlot, Francis Bach. Multi-task Regression using Minimal Penalties. JMLR 13(Sep):2773-2812, 2012.
    [journal]   [pdf]   [Hal]   [arXiv]   [matlab code]   [slides Cambridge november 2011]   [slides CIRM december 2013]
  15. (2011) Sylvain Arlot, Peter L. Bartlett. Margin adaptive model selection in statistical learning. Bernoulli, Vol. 17, No 2, 687--713. DOI: 10.3150/10-BEJ288.
    [doi]   [journal]   [Euclid]   [Hal]   [arXiv]
  16. (2010) Sylvain Arlot, Alain Celisse. Segmentation of the mean of heteroscedastic data via cross-validation. Statistics and Computing, 1--20. DOI: 10.1007/s11222-010-9196-x.
    [journal]   [pdf]   [supplementary material, pdf]   [short version, 41e Journées de Statistique, SFDS]   [Hal]   [arXiv]   [matlab code]   [slides Berlin april 2009]
  17. (2010) Sylvain Arlot, Alain Celisse. A survey of cross-validation procedures for model selection. Statistics Surveys, 4, (2010), 40-79 (electronic). DOI: 10.1214/09-SS054.
    [journal]   [Hal]   [arXiv]   [slides IHP april 2015]
  18. (2010) Sylvain Arlot, Gilles Blanchard, Etienne Roquain. Some nonasymptotic results on resampling in high dimension, I: Confidence regions. The Annals of Statistics, Vol. 38, No. 1, 51-82.
    [journal]   [Euclid]   [Hal]   [arXiv]   [matlab code]   [slides Rennes october 2010]
  19. (2010) Sylvain Arlot, Gilles Blanchard, Etienne Roquain. Some nonasymptotic results on resampling in high dimension, II: Multiple tests. The Annals of Statistics, Vol. 38, No. 1, 83-99.
    [journal]   [Euclid]   [Hal]   [arXiv]   [matlab code]   [slides Rennes october 2010]
  20. (2009) Sylvain Arlot, Francis Bach. Data-driven calibration of linear estimators with minimal penalties. Advances in Neural Information Processing Systems (NIPS) 22, 46-54.
    [proceedings]   [pdf]   [supplementary material]   [slide]   [poster]   [slides Cambridge november 2011]   [slides CIRM december 2013]     -     Long version (preprint): [Hal]   [arXiv]
  21. (2009) Sylvain Arlot. Model selection by resampling penalization. Electronic Journal of Statistics, 3, (2009), 557-624 (electronic). DOI: 10.1214/08-EJS196.
    [journal]   [appendix, pdf]   [Hal]   [matlab code]   [slides EMS 2009]
    See also a previous 2007 preprint, short version of this paper:   [Hal]   [arXiv]
  22. (2009) Josselin Desmars, Sylvain Arlot, Jean-Eudes Arlot, Valery Lainey, Alain Vienne. Estimating the accuracy of satellite ephemerides using the bootstrap method. Astronomy and Astrophysics, 499, 321--330.
    [journal]   [abstract]   [slides Pisa march 2012]
  23. (2009) Sylvain Arlot, Pascal Massart. Data-driven calibration of penalties for least-squares regression. Journal of Machine Learning Research. 10(Feb):245--279, 2009
    [journal]   [pdf]   [Hal]   [arXiv]   [slides Berlin april 2009]   [slides CIRM december 2013]
  24. (2007) Sylvain Arlot, Gilles Blanchard, Etienne Roquain. Resampling based confidence regions and multiple tests for a correlated random vector. In COLT 2007, volume 4539 of Lecture Notes in Artificial Intelligence, pp. 127-141. Springer, Berlin, 2007.
    [proceedings]   [Hal]   [arXiv]   [slides Rennes october 2010]
Other papers (non peer-reviewed) -- Autres articles (publiés sans processus de revue)
  1. (2018) Sylvain Arlot, Stefano Marmi, Carlos Matheus et Nigel G. Yoccoz. Dynamique de populations de petits mammifères, saisonnalité et attracteur de Hénon -- comment une question d'écologue a pu intéresser Jean-Christophe. In Sylvain Crovisier, Pierre Berger, Patrice Le Calvez et Carlos Matheus, éditeurs, La Gazette des Mathématiciens, numéro spécial en l'honneur de Jean-Christophe Yoccoz, avril 2018.
  2. (2010) Sylvain Arlot, Francis Bach. Data-driven penalties for linear estimators selection. In Oberwolfach Reports. Vol. 7, no. 1, report No.16/2010. Workshop "Modern Nonparametric Statistics: Going Beyond Asymptotic Minimax".
    [pdf]   [workshop page]   [complete report, pdf]
  3. (2008) Sylvain Arlot. Comment choisir un modèle? Le mensuel de l'université.
    [article]
  4. (2007) Sylvain Arlot. V-fold penalization: an alternative to V-fold cross-validation. Oberwolfach Report 50-2007, workshop "Reassessing the paradigms of statistical model-building".
    [pdf]   [workshop page]   [complete report, pdf]
My "habilitation" and Ph.D. theses -- Mon mémoire d'habilitation à diriger des recherches et ma thèse
  1. (2014) Sylvain Arlot. Contributions to statistical learning theory: estimator selection and change-point detection, Université Paris Diderot. Habilitation thesis.
    [pdf]   [TEL]   [slides]
  2. (2007) Sylvain Arlot. Resampling and Model selection, Université Paris-Sud. PhD thesis.
    [abstract]   [pdf]   [TEL]   [slides]   [dedicated webpage]
Unpublished texts -- Textes non publiés
  1. (2015) Rémi Lajugie, Piotr Bojanowski, Sylvain Arlot and Francis Bach. Semidefinite and Spectral Relaxations for Multi-Label Classification Preprint.
    [Hal]   [arXiv]   [matlab code]
  2. (2010) Sylvain Arlot. Choosing a penalty for model selection in heteroscedastic regression. Preprint.
    [Hal]   [arXiv]   [slides EMS 2009]
  3. (2008) Sylvain Arlot. V-fold cross-validation improved: V-fold penalization. Preprint.
    [Hal]   [appendix, pdf]   [arXiv]   [slides IHES march 2013]   [slides Toulouse june 2008]
  4. (2004) Sylvain Arlot. Étude d'un modèle de dynamique des populations. Master 2 report, directed by J.-C. Yoccoz (in French).
    [Hal (report + videos in appendix)]   [arXiv]   Report with full resolution figures [ps.gz]   [slides Siena november 2009]   [dedicated webpage]

Software -- Logiciels


Last update: 2022/11/28