Ambroise Heurtebise

I received my PhD in machine learning and statistical signal processing from Inria Saclay (Université Paris-Saclay), working in the MIND team under the supervision of Alexandre Gramfort and Pierre Ablin.

Before my PhD, I completed a Master’s degree in Data Science at the Institut Polytechnique de Paris, following a four-year undergraduate and master-level program in Mathematics at Université Paris-Dauphine.

Email  /  GitHub  /  Google Scholar  /  LinkedIn

Research

My research focuses on multi-view learning, independent component analysis (ICA), and causal discovery, as well as on learning useful representations of brain activity.

Multi-View Causal Discovery without Non-Gaussianity: Identifiability and Algorithms

We learn a causal ordering (Directed Acyclic Graph) between random variables collected from different environments, such as brain regions.

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

We extend the time variability of MVICAD by allowing sources to also differ by time dilations, which is for example useful when a subject performs an auditory task.

MVICAD: Multi-View Independent Component Analysis with Delays

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.

Personal work

I developed a website dedicated to the game Yams (also known as Yahtzee), featuring score sheets, statistics, and trophies. The website is available at https://myams.fr. A secret code is required to sign up, so feel free to contact me if you would like to get access, or if you have any questions or suggestions about the project.

Design and source code from Jon Barron's website