Omar Chehab

As of June 2025, I joined the Machine Learning Department at Carnegie Mellon University, as a Postdoctoral Research Associate in Pradeep Ravikumar’s team. My previous postdoc was in the Statistics Department of CREST-ENSAE, working with Anna Korba’s team.

In November 2023, I completed my PhD in mathematical computer science at Inria, where I was advised by Aapo Hyvärinen and Alexandre Gramfort. My graduate studies led to a Master’s in engineering from ENSTA Paris and in Applied Maths, Vision and Learning (MVA) from ENS Paris-Saclay.

Email  /  GitHub  /  Google Scholar  /  LinkedIn

Research

My research is in machine learning, particularly on efficient algorithms for estimating and sampling from (energy-based) statistical models, as well as on learning useful representations of brain activity.

Sampling from Energy-Based Statistical Models

Sampling from multi-modal distributions with polynomial query complexity in fixed dimension via reverse diffusion

Time-reversed diffusions are state-of-the-art for sampling multi-modal distributions, but they rely on score estimates. We analyze how estimation errors affect the final samples.

Provable Convergence and Limitations of Geometric Tempering for Langevin Dynamics

Annealed MCMC tries to approximate a prescribed path of distributions. We show that the popular geometric mean path with a Gaussian has unfavorable geometry. Presented at the Yale workshop on sampling.

A Practical Diffusion Path for Sampling

Time-reversed diffusions are state-of-the-art in sampling but rely on score estimates. We aim to reduce their variance.

Estimating Energy-Based Statistical Models

Density Ratio Estimation with Conditional Probability Paths

A density ratio can be obtained by integrating the time score of a probability path. We present an efficient way to estimate the time score.

Provable benefits of annealing for estimating normalizing constants: Importance Sampling, Noise-Contrastive Estimation, and beyond

Annealed Importance Sampling uses a prescribed path of distributions to compute an estimate of a normalizing constant. We quantify how the choice of path impacts the estimation error.

The Optimal Noise in Noise-Contrastive Learning Is Not What You Think

NCE estimates the data density by minimizing a binary classification loss, between data and noise samples. We find the optimal noise distribution that minimizes the estimation error.

Learning Representations of Brain Activity

Identifiable Multi-View Causal Discovery Without Non-Gaussianity

We learn a causal ordering (Directed Acyclic Graph) between random variables collected from different environments.

MVICAD2: Multi-View Independent Component Analysis with Delays and Dilations

Independent Component Analysis (ICA) is a popular algorithm for learning a representation of data. We propose a version that handles data collected from different contexts, and whose representations differ only by temporal delays or dilations.

Deep Recurrent Encoder: an end-to-end network to model magnetoencephalography at scale

We compare different models for predicting the brain’s response to external stimuli. Our model, based on a deep neural network, is more accurate and interpretable.

Learning with self-supervision on EEG data

We learn rich representations of EEG brain activity using a self-supervised loss.

Uncovering the structure of clinical EEG signals with self-supervised learning

We learn rich representations of EEG brain activity using a self-supervised loss.

A mean-field approach to the dynamics of networks of complex neurons, from nonlinear Integrate-and-Fire to Hodgkin–Huxley models

Our theory predicts the average behavior of neuronal populations that fire asynchronously.

Talks

* 06/2025, Risteski Group Seminar, Carnegie Mellon, USA
* 04/2025, Department of Statistics and Data Science Seminar, NUS, Singapore
* 03/2025, DATA group Seminar, Laboratoire Jean Kuntzmann, France
* 03/2025, Anima AI + Science Lab Seminar, Caltech, USA
* 12/2024, Kempner Institute, Harvard, USA
* 10/2024, Signals, Information, and Algorithms Laboratory Seminar, MIT, USA
* 10/2024, Uncertainty Quantification Group Seminar, MIT, USA
* 03/2024, Statistical Mathematics Group Seminar, RIKEN, Japan
* 03/2023, Self-Supervised Learning Reading Group, Vector Institute, Canada
* 02/2020 First International Workshop on Nonlinear ICA, Inria, France

Teaching

I was Teacher's Assistant for the following Masters courses.

Optimization for Data Science - Institut Polytechnique de Paris (2021-2023)
Professors: Alexandre Gramfort, Pierre Ablin
Advanced Machine Learning - CentraleSupelec, Universite Paris-Saclay (2020-2022)
Professors: Emilie Chouzenoux, Frederic Pascal
Optimization - CentraleSupelec, Universite Paris-Saclay (2020-2021)
Professors: Jean-Christophe Pesquet, Sorin Olaru, Stephane Font

Service

I review submissions to the following machine learning conferences: NeurIPS, ICML, ICLR and AISTATS.

I occasionally review submissions to the following journals: the Journal of Machine Learning Research (JMLR), the Annals of the Institute of Statistical Mathematics (AISM), and Mathematical Methods of Statistics.

Design and source code from Jon Barron's website