Publications

Machine Learning

NEO: Non Equilibrium Sampling on the Orbit of a Deterministic Transform

Sampling from a complex distribution $\pi$ and approximating its intractable normalizing constant Z are challenging problems. In this paper, a novel family of importance samplers (IS) and Markov chain Monte Carlo (MCMC) samplers is derived. Given an…

Achille Thin, Yazid Janati, Sylvain Le Corff, Charles Ollion, Arnaud Doucet, Alain Durmus, Eric Moulines, Christian Robert

35th International Conference on Neural Information Processing Systems - 2021

Machine Learning

KFC: A clusterwise supervised learning procedure based on the aggregation of distances

Nowadays, many machine learning procedures are available on the shelves and may be used easily to calibrate predictive models on supervised data. However, when the input data consists of more than one unknown cluster, linked to different underlying…

Sothea Has, Aurélie Fisher, Mathilde Mougeot

Journal of Statistical Computation and Simulation - 2021

Statistics Theory

On bayesian estimation and proximity operators

There are two major routes to address the ubiquitous family of inverse problems appearing in signal and image processing, such as denoising or deblurring. A first route relies on Bayesian modeling, where prior probabilities are used to embody models…

Rémi Gribonval, Mila Nikolova

Applied and Computational Harmonic Analysis - 2021

Probability

The effect of discretization on the mean geometry of a 2D random field

The study of the geometry of excursion sets of 2D random fields is a question of interest from both the theoretical and the applied viewpoints. In this paper we are interested in the relationship between the perimeter (resp. the total curvature,…

Hermine Biermé, Agnès Desolneux

Annales Henri Lebesgue - 2021

Statistics

Analysis of Least square estimator for simple Linear Regression with a uniform distribution error

We study the least square estimator, in the framework of simple linear regression, when the deviance term ϵ with respect to the linear model is modeled by a uniform distribution. In particular, we give the law of this estimator, and prove some…

M R Jlibene, S Taoufik, S Benjelloun

Ocean, Atmosphere

Can we improve the realism of gravity wave parameterizations by imposing sources at all altitudes in the atmosphere?

A multiwave non-orographic gravity waves (GWs) scheme is adapted to represent waves of small intrinsic phase speed, inertial waves, and wave emission from all altitudes. This last change removes the launching altitude parameter, an arbitrary…

B Ribstein, C Millet, F Lott, A de La Cámara

Journal of Advances in Modeling Earth Systems - 2021

Methodology

Maximum entropy methods for texture synthesis: theory and practice

Recent years have seen the rise of convolutional neural network techniques in exemplar-based image synthesis. These methods often rely on the minimization of some variational formulation on the image space for which the minimizers are assumed to be…

Valentin de Bortoli, Agnès Desolneux, Alain Durmus, Bruno Galerne, Arthur Leclaire

SIAM Journal on Mathematics of Data Science - 2021

Engineering Sciences

A novel hybrid drone for multi-propose aerial transportation and its conceptual optimization based on surrogate approach

Abstract In recent years the Unmanned Aerial Vehicles (UAVs) are increasingly adopted in various applications of the aerial transportation, meanwhile both the two main classifications, the fixed wing and multi-rotors drones, are limited by their own…

Hao Yue, Hicham Medromi, Hua Ding, David Bassir

Journal of Physics: Conference Series - 2021

Engineering Sciences

A Refined Theory for Bending Vibratory Analysis of Thick Functionally Graded Beams

A refined beam theory that takes the thickness-stretching into account is presented in this study for the bending vibratory behavior analysis of thick functionally graded (FG) beams. In this theory, the number of unknowns is reduced to four instead…

Youssef Boutahar, Nadhir Lebaal, David Bassir

Mathematics - 2021

Probability

Piecewise Deterministic Markov Processes and their invariant measure

Piecewise Deterministic Markov Processes (PDMPs) are studied in a general framework. First, different constructions are proven to be equivalent. Second, we introduce a coupling between two PDMPs following the same differential flow which implies…

Alain Durmus, Arnaud Guillin, Pierre Monmarché

Annales de l'Institut Henri Poincaré (B) Probabilités et Statistiques - 2021