Summer school

S. Chrétien,  Optimization Methods for Machine Learning, EPAT 2010, Mai 2-7, Baie de Somme, Hotel Cap Hornu.


Conferences 


S. Chretien and S. Darses, An elementary approach to the problem of column selection in a rectangular matrix, The 2nd IMA Conference on the Mathematical Challenges of Big Data, Mary Ward House, London on Thursday 1st – Friday 2nd December 2016.

S. Chretien, N. Herr, J.-M. Nicod, C. Varnier - Scheduling independent parallel machines with convex programming, PGMO Days 2016, 8-9 November 2016.

S. Chretien, P. Clarkson, A. Forbes - A convex relaxation of the optimal sensor placement problem in power grids, PGMO Days 2016, 8-9 November 2016.

S. Chretien, P. Clarkson and A. Forbes, Optimal sensor placement in Smart Grids: an approach based on semi-definite relaxation and sensitivity analysis, ENBIS 201611 – 15 September 2016, Sheffield, UK. 

S. Chretien, P. Harris and R. Tawil, Total Variation minimization for Compressed Sensing with ``smoothly'' varying covariates, IEEE - CSE 2016, August 24-26 2016, Paris.  

S. Chretien, The Von Neumann trace inequality for tensors (Joint work with T. Wei), ILAS 2016, July 11-15 2016, KU Leuven, Belgium. 

Stéphane Chrétien, Nathalie Herr, Jean-Marc Nicod and Christophe Varnier, Post-Prognostics Decision for the Commitment of Fuel Cell System, 3rd IFAC Workshop on Advanced Maintenance Engineering, Service and Technology, AMEST-16, Biarritz (France), October 19-21. 

S. Chretien and T. Wei, A Lagrangian view point on robust PCA, Learning Theory, CIRM, (France), 1-6 February 2016.

S. Chretien, J. Coupey, J.-M. Nicod, C. Varnier, Biosolver: a VRPTW solver for the nurses tour scheduling problem with hard time constraints, Annual workshop of the EURO working group on Vehicle Routing and Logistics optimization (VeRoLog), 6-8 Jun 2016 Nantes (France)

S. Chretien, T. Wei and B. Al-Sarray, Joint estimation and model order selection for one dimensional ARMA models via convex optimization,  8th International Conference of the ERCIM WG on Computational and Methodological Statistics (CMStatistics 2015), Senate House, University of London, UK, 12-14 December 2015.

S. Chrétien and C. Guyeux, Using the Lasso for gene selection in bladder cancer data 8th International Conference of the ERCIM WG on Computational and Methodological Statistics (CMStatistics 2015), Senate House, University of London, UK, 12-14 December 2015.

S. Chrétien and T. Wei, Convex recovery of tensors using nuclear norm penalization,
12th International Conference on Latent Variable Analysis and Signal Separation, LVA 2015. LNCS, LVA-ICA 2015, E. Vincent, A. Yeredor, Z. Koldovský and P. Tichavský (editors)ArXiv.

S. Chretien, N. Herr, J.-M. Nicod, C. Varnier, A Post-Prognostics Decision Approach to Optimize the Commitment of Fuel Cell Systems in Stationary Applications, PHM 2015, Austin, TX, USA. 

S. Chrétien, C. Guyeux, B. Conesa, R. Delage-Mourroux, M. Jouvenot, P. Huetz and F. Descotes, Estimating features with missing values and outliers: a Bregman-proximal point algorithm for robust Non-negative Matrix Factorization with application to gene expression analysis, LNCS ISBRA 2015, to appear. ArXiv.

S. Chrétien, Estimation of Gaussian Mixture Models via mixted nuclear/$\ell_\infty$/$\ell_1$-norm penalization, Methodological advances in Statistics related to Big Data, Castro Urdiales, june 8-12, 2015.

S. Chrétien and T. Wei, Von Neumann's trace inequality for tensors, TMA 2014, slides.

S. Chrétien, Estimation of Gaussian Mixture Models via mixted nuclear/$\ell_\infty$/$\ell_1$-norm penalization, ERCIM 2014, Pisa, 6-8 december 2014. Slides

S. Chrétien, N. Herr, J.-M. Nicod, C. Varnier, Scheduling independent parallel machines with convex programming, PGMO-COPI 2014, Ecole Polytechnique, Palaiseau, France, october 27-31 2014. Slides

S. Chrétien, Convex optimization methods for ARMA time series with trend and seasonality, Time Series Econometrics and Finance, Besancon, France, 5-6 mai 2014.

S. Chrétien, Mixture model for designs in high dimensional regression and the LASSO, European Meeting of Statisticians, Budapest, Hungary, 20-25 July 2013.

S. Chrétien, L. Toch, J.-M. Nicod, L. Philippe and V. Sonigo, Job Scheduling Using successive Linear Programming Approximations of a Sparse Model,  18th International European Conference on Parallel and Distributed Computing, Rhodes Island, Greece, August 27th-31st, 2012.

S. Chrétien, Prediction bounds for the LASSO without incoherence, 8th World Congress in Probability and Statistics, Istanbul, July 9-14, 2012.

S. Chretien and S. Darses, Sparse recovery with unknown variance: a LASSO-type approach, Imaging, Communications and Finance: Stochastic Modeling of Real-world Problems, Conference in honor of Lawrence A. Shepp, June 24-25, 2011, Columbia University, New York, NY. (Poster)

Eric Bernard, J.-M. Friedt, G. Martin, L. Moreau, D. Laffly, S. Chretien, C. Marlin, M. Griselin,
Automated high resolution digital image acquisition and processing applied to polar glaciers, Polar Worlds, Paris, France,  26-28 january 2011. Abstract

S. Chretien and J.-P. Ortega, Spectral sparsity and the modeling of high dimensional heteroscedastic phenomena, 4th CSDA International Conference on Computational and Financial Econometrics (CFE 10) Senate House, University of London, UK 10-12 December 2010.

S. Chretien, l1-penalized maximum likelihood estimation for sparse polytomic regression models, EMS 2010, August 17-22, University of Piraeus.

S. Chretien, Lagrangian relaxation for the Compressed Sensing problem, SMAI 09, Mai 2009, La Colle sur Loup. Slides

 S. Chretien, The two-stage l1 relaxation for the Compressed Sensing problem, SPARS 09, April 2009, Saint Malo. (Poster)

S. Chretien and J.-P. Ortega, Decay of correlation for a discretized differential delay equation, BIRS 07w5068 Geometric Mechanics: Continuous and discrete, finite and infinite dimensional, Auguts 2007, Banff.

S. Chretien, Tighter relaxations for the sparsest recovery problem, ICIAM 07, Zurich. Slides Proceedings (non free)

S. Chretien and J.-P. Ortega, Decay of correlation for a discretized differential delay equation, IMWSM 2007, Sevilla. Slides

S. Chretien, Relaxed MCMC for random fields with quadratic Hamiltonians over binaries. 25 th European Meeting of Statisticians, Oslo, 2005.Slides.

    S. Chretien and F. Corset, Une borne inferieure pour les temps d'inspection en maintenance. Journees de la Societe Francaise de Statistiques 2005.

    S. Chretien and F. Corset, A lower bound on inspection time for complex systems with Weibull transitions. COMPSTAT 2004 (Prague), 799--806, Physica, Heidelberg, 2004. Proceedings (non free).(Poster)

    S. Chretien and F. Corset, A lower bound on inspection time for complex systems with Weibull transitions. MMR 2004.

    S. Chretien and F. Corset, Reconstruction d'images binaires par estimation moindres carres et optimisation de valeur propre. Journees de la Societe Francaise de Statistiques 2004.

    S. Chretien and F. Corset, Least squares reconstruction of binary images using eigenvalue optimization. COMPSTAT 2002 (Berlin), 419--424, Physica, Heidelberg, 2002. Proceedings (non free).

    C. Biernacki and S. Chretien, Degeneracy in the likelihood approach to univariate gaussian mixture estimation with EM. ASMDA 2001 (Compiegne), 206--212, 2001. Proceedings (free access).

    S. Chretien, EM-type algorithm for maximum likelihood estimation and application to medical imaging, (Invited Speaker), journees de la Societe Francaise de Statistiques, Fes, Maroc, May 2000. Abstract , Slides.

    S. Chretien, A.O. Hero and R. Piramuthu, Statistical Proximal Point Methods for Image Reconstruction, SIAM Annual Meeting, Puerto Rico, 10--14 juillet 2000. Link, Abstract , Slides.

    S. Chretien, Generalized proximal point algorithms and bundle implementation, Conference groupe MODE de la SMAI, Orleans, France, mars 1999, Link .

    S. Chretien and A. O. Hero, Acceleration of the EM algorithm via proximal point iterations, IEEE International Symposium on Information Theory, Boston, August 1998. Paper (non free).

    S. Chretien and I. Dologlou, Successive projection-like algorithms for signal approximation/zero error modelling, Proc. IEEE international conference on acoustics, speech and signal processing, Detroit, USA, (1995), 1240-1243. Paper (non free)

    S. Chretien and I. Dologlou, Image compression based on reduced rank approximation and adaptive error modelling, Proc. European Signal Processing Conference, Edinburgh, UK, (1994), 592-595.

Talks  



The LASSO when the design is not coherent, Quantitative Methods Seminar, St Gallen, Switzerland, March 13 2017. 

An elementary approach to Robust PCA, Seminaire de Statistiques, Universite de Bourgogne Franche-Comte, September 2016, Dijon, France.

Optimal sensor placement in Smart Grids: an approach based on semi-definite relaxation and sensitivity analysis, Universite de Franche-Comte, Probability and Statistics Seminar, 11 September 2016, Besancon, France. 

Some low rank estimation problems in Times Series and Clustering, UCL, London, July 7, 2016. 

Estimation of Gaussian Mixture Models via mixted nuclear/$\ell_\infty$/$\ell_1$-norm penalization, Statistics and Probability seminar, Lille 1, may 24, 2016.

Survey on sparse recovery, Seminar, University of Surrey, Guildford, May 5 2016. 

Low rank tensor recovery via convex optimization, Statistics seminar, LJK Grenoble, april 23 2015. Slides.

Clustering par optimisation convexe, Statistics seminar, ENS Rennes, april 10 2015. Slides.

A new criterion for design matrices in the LASSO, Probabilty seminar, Université du Luxembourg, february 26 2015. Slides.

Le LASSO pour des matrices de design non necessairement incoherentes, Séminaire Signal-Apprentissage. Université Aix-Marseille, march 15 2012. Slides and Video

Le Lasso avec variance inconnue. Séminaire de Statistique INRIA Lille-Nord Europe, mai 31 2011, Villeneuve d'Asq. Slides

Le Lasso avec variance inconnue. Séminaire de Statistique Université Montpellier 2, mai 9 2011, Montpellier. Slides

Le Lasso avec variance inconnue. Séminaire de Statistique Université Paul Sabatier, march 29 2011, Toulouse. 

Lasso et extensions, ou le compressed sensing en milieu bruité, Groupe de Travail Compressed Sensing, Institut Henri Poincaré, March 11 2009. Paris. 

Lagrangian relaxation of the compressed sensing problem, Statistics seminar, Boston University, November 4 2008. Boston MA   

    Relaxation lagrangienne du probleme de Compressed Sensing. Rencontres Probabilites-Statistiques Besancon-Dijon, January 2008. Dijon.

    Introduction au Compressed Sensing. Seminaire de Probabilite et Statistiques, November 2007. Besancon.

    Une generalisation du modele de Diggle, Tawn et Moyeed en geostatistique, (with D. Pleydell) Seminaire de Probabilite et Statistiques, Universite de Lille 1, mai 2006. Abstract.

    A Stochastic Bundle Method, Seminaire SMG, Han sur Lesse (Belgique), Avril 2000 Slides.

    A Short Introduction to Bundle Methods, Seminaire du Service de Mathematiques de la Gestion, Universite Libre de Bruxelles, Fevrier 2000 Slides.

    Semidefinite Programming and Combinatorial Optimization, Seminaire SMG, Borzee (Belgique), mai 1999 Slides.

    Approches EM et Kullback-proximale pour la maximisation de vraisemblance et application aux melanges, Seminaire IS2, INRIA Rhone Alpes, Slides.

    Approches EM et Kullback-proximale pour la maximisation de vraisemblance, Seminaire LMC Universite Joseph Fourier Grenoble, decembre 1998.


Back to homepage     Math Home Page