Research topics :

Keywords: Machine Learning &  Hidden Markov Chains ; Sequential Monte Carlo for on-line estimation; Stochastic volatility models

Book:

L'apprentissage profond (2018), a French translation of Deep Learning (I. Goodfellow,  Y. Bengio, A. Courville), Florent Massot & Quantmetry. "The first book on deep learning translated using deep learning". Editor and lead traducer.  With F. Navarro, S. El Kolei, B. Guedj, C. Chesneau, N. Bousquet et al. 

Prints:

• Investigating swimming technical skills by a double partition clustering of multivariate functional data allowing for dimension selection, with A. Bouvet and M. Marbac, 2023 : ArXiv  (accepted in Annals of Applied Statistics)

• Parametric estimation of hidden Markov models by least squares type estimation and deconvolution (with C. Chesneau et F. Navarro), 2022: Statistical Papers

• Contrast estimation for noisy obervations of diffusion processes via closed-form density expansions (with F. Navarro), 2021: Statistical Inference for Stochastic processes

• Nonparametric estimation in a regression model with additive and multiplicative noise (with C. Chesneau, J. Kou, F. Navarro), 2020: Journal of Computational and Applied Mathematics, Vol. 380

•  Adaptive density estimation on bounded domains (with K. Bertin and N. Klutchnikoff), 2019 :  Annales de l'Institut Henri Poincaré (B) , ArXiv

•  Analysis, detection and correction of misspecified discrete time state space models, (with F. Patras), 2018 : Journal Of Computational and Applied Mathematics , ArXiv

• Parametric inference of autoregressive heteroscedastic models with error in variables (with F. Pelgrin), 2017 :  Statistics & Probability Letters,  ArXiv

•  Parametric estimation of hidden stochastic models by deconvolution and contrast minimization: application to the stochastic volatility models, 2013 : Metrika, ArXiv.

Pre-prints:

• Estimation of the Order of Non-Parametric Hidden Markov Models using the Singular Values of an Integral Operator,  with M. Du Roy de Chaumaray, M.P. Etienne  and M. Marbac : ArXiv

• Kernel Smoothing Conditional Particle filter with Ancestor Sampling, with F. Navarro.

Working-Papers:

• Propagation of initial error on the parameters for linear and Gaussian state-space models : ArXiv

• Géométrie des matrices de covariance pour le traitement des signaux radars : Hal

PhD Supervision

• September 2022-present: Antoine Bouvet Analyzing inertial swimming data for automatically monitoring athlete's activity during training. Jointly supervised with M. Marbac (CREST-Ensai) and Nicolas Bideau (M2S)

PhD Thesis :

 Estimation des modèles à volatilité stochastique par filtrage et déconvolution : Manuscrit