Bertrand MICHEL - Publications and Softwares


  • An introduction to Topological Data Analysis: fundamental and practical aspects for data scientists, with F. Chazal. Preprint

  • Estimating the Reach of a Manifold, with E. Aamari, J. Kim, F. Chazal, A. Rinaldo and L. Wasserman. Preprint

  • Detecting and modeling worst-case dependence structures between random inputs of computational reliability models. Nazih Benoumechiara, Bertrand Michel, Philippe Saint-Pierre, Nicolas Bousquet. Preprint


  • Activity recognition from stride detection: a machine learning approach based on geometric patterns and trajectory reconstruction, with B. Beaufils, F. Chazal and M. Grelet.. Conference IPIN 2018 Preprint

  • Statistical analysis and parameter selection for Mapper, with M. Carrière and S. Oudot. Journal of Machine Learning Research, Vol 19 N12 Preprint

  • Stride detection for pedestrian trajectory reconstruction: a machine learning approach based on geometric patterns, with B. Beaufils, F. Chazal and M. Grelet. Conference IPIN 2017 Preprint.

  • Data driven estimation of Laplace-Beltrami operator. F. Chazal, I. Giulini, B. Michel. The Thirtieth Annual Conference on Neural Information Processing Systems (NIPS 2016). Preprint

  • Rates of convergence for robust geometric inference, with F. Chazal and P. Massart. Electron. J. Statist, Volume 10, 2243-2286.

  • Grouped variable importance with random forests and application to multivariate functional data analysis, with B. Gregorutti and P. Saint-Pierre Accepted for pubication in CSDA

  • Improved rates for Wasserstein deconvolution with ordinary smooth error in dimension one, with J. Dedecker and A. Fischer Accepted for publication in EJS.

  • Minimax rates of convergence for Wasserstein deconvolution with supersmooth errors in any dimension, with J. Dedecker. Accepted for publication in JMVA.

  • Adaptive density estimation using finite Gaussian mixtures, with C. Maugis.
    ESAIM P&S 2013, Volume 17, pp 698 - 724.

  • Data-driven penalty calibration : A case study for Gaussian model selection, with C. Maugis.
    ESAIM P&S 2011, Vol 15, p.320-339, Preprint

  • A non asymptotic penalized criterion for Gaussian mixture model selection, with C. Maugis. ESAIM P&S] 2011, Vol 15, p.41-68.


Thesis and Habilitation thesis

  • Habilitation thesis : A Statistical Approach to Topological Data Analysis Manuscrit.

  • Thesis : Modélisation de la production d'hydrocarbures dans un bassin pétrolier, 2008, Université Paris Sud Orsay : Manuscript.