Since 2017, an AI has been defeating the best Go experts, despite the game being particularly challenging. Such “super intelligence” is rare, but it could also emerge in fundamental mathematics.
Note
This post was originally published in French as part of my scientific chronicle …
Is unbiased data important to build health AI? Yes!
Can there be unbiased data? No!
The notion of bias depends on the intended use.
In medicine, we have seen the importance of tuning devices and decisions for the …
Note
For me, 2024 was full of back and forth between research, software, and connecting these to society. Here, I lay out some highlights on AI and society, as well as research and software, around tabular AI and language models.
As 2025 starts, I’m looking back on 2024. It …
Improving conversational artificial intelligences or simpler prediction engines involves overcoming biases, that is, going beyond the limits of data. But the notion of bias is subtle, as it depends on the goals.
Note
This post was originally published in French as part of my …
Despite their apparent intelligence, conversational artificial intelligences often lack logic. The debate rages on: do they reason or do they recite snatches of text memorized on the Internet?
Note
This post was originally published in French as part of my scientific chronicle in Les …
Note
Foundation models, pretrained and readily usable for many downstream tasks, have changed the way we process text, images, and sound. Can we achieve similar breakthroughs for tables? Here I explain why with “CARTE”, we’ve made significant headway.
Contents
English summary
I have been appointed to the government-level panel of experts on AI, to set the national vision and strategy in France.
J’ai l’honneur d’être nommé au comité de l’intelligence artificielle du gouvernement Français.
La mission qui nous est confiée d’éclairer l’action publique …
A retrospective on last year (2022): I embarked on a new scientific adventure, assembling a team focused on developing machine learning for health and social science. The team has existed for almost a year, and the vision is nice shaping up. Let me share with you illustrations of where we …
Broad decoding models that can specialize to discriminate closely-related mental process with limited data
TL;DR
Decoding models can help isolating which mental processes are implied by the activation of given brain structures. But to support a broad conclusion, they must be trained on many studies, a difficult problem given …
The year 2020 has undoubtedly been interesting: the covid19 pandemic stroke while I was on a work sabbatical in Montréal, at the MNI and the MILA, and it pushed further my interest in machine learning for health-care. My highlights this year revolve around basic and applied data-science for health.
Highlights …
Note
A simple survey asking authors of two leading machine-learning conferences a few quantitative questions on their experimental procedures.
How do machine-learning researchers run their empirical validation? In the context of a push for improved reproducibility and benchmarking, this question is important to develop new tools for model comparison. We …
My current research spans wide: from brain sciences to core data science. My overall interest is to build methodology drawing insights from data for questions that have often been addressed qualitatively. If I can highlight a few publications from 2019 [1], the common thread would be computational statistics, from dirty …
Note
Given two set of observations, are they drawn from the same distribution? Our paper Comparing distributions: l1 geometry improves kernel two-sample testing at the NeurIPS 2019 conference revisits this classic statistical problem known as “two-sample testing”.
This post explains the context and the paper with a bit of hand …
From a scientific perspective, 2018 [1] was once again extremely exciting thank to awesome collaborators (at Inria, with DirtyData, and our local scikit-learn team). Rather than going over everything that we did in 2018, I would like to give a few highlights: We published major work using machine learning to …
In my opinion the scientific highlights of 2017 for my team were on multivariate predictive analysis for brain imaging: a brain decoder more efficient and faster than alternatives, improvement clinical predictions by predicting jointly multiple traits of subjects, decoding based on the raw time-series of brain activity, and a personnal …
Year 2016 has been productive for science in my team. Here are some personal highlights: bridging artificial intelligence tools to human cognition, markers of neuropsychiatric conditions from brain activity at rest, algorithmic speedups for matrix factorization on huge datasets…
Eickenberg et …
My research group is looking to fill a post-doc position on learning biomarkers from functional connectivity.
The challenge is to use resting-state fMRI at the level of a population to understand how intrinsic functional connectivity captures pathologies and other cognitive phenotypes. Rest fMRI is a promising tool for …
Christophe Pradal, Hans Peter Langtangen, and myself recently edited a version of the Journal of Computational Science on scientific software, in particular those written in Python. We wrote an editorial defending writing and publishing open source scientific software that I wish to summarize here. The full text preprint is openly …
Lately, I have been a mood of scientific scepticism: I have the feeling that the worldwide academic system is more and more failing to produce useful research. Christophe Lalanne’s twitter feed lead me to an interesting article in a non-mainstream journal: A farewell to Bonferroni: the problems of low …
At the request of a friend, I am putting up some of the posters that I recently presented at conferences.
Large-scale functional-connectivity graphical models for individual subjects using population prior.
This is a poster for our NIPS work
Multi-subject dictionary learning to segment an atlas of brain spontaneous activity …
