Postdoctoral researcher
L’Ecole normale supérieure - PSL
franco.pellegr [at] gmail.com
PhD in Condensed Matter physics from SISSA, Trieste, Italy
Theory of neural network learning dynamics.
We want to describe the general mechanism that makes neural networks so effective in solving real life problems. We aim to use methods from statistical physics to build a model describing the dynamics of neural network parameters during training. We hope this insight will allow us to develop new training algorithms leading to improved networks.
Postdoctoral researcher
L’Ecole normale supérieure - PSL
misaki.ozawa2045 [at] gmail.com
PhD University of Tsukuba
Multiscale physics and wavelet transform.
In physics, multiscale phenomena are captured by the renormalization group. Wavelet transform is useful in analyzing multiscale behaviors. We investigate the relation between the renormalization group and the wavelet transform. Then we wish to obtain insight into how features are extracted hierarchically in neural networks.
Postdoctoral researcher
L’Ecole normale supérieure - PSL
Tanguy.marchand [at] ens.fr
Simulating physical stochastic processes using Machine Learning.
Physical stochastic processes such as turbulence, astrophysical maps and so on provide a wide range of non-Gaussian processes. My research is to develop new Machine Learning tools to better analyze and reproduce them.
Postdoctoral researcher
L'Ecole normale supérieure - PSL
georgialoukatou [at] gmail.com
PhD, École Normale Supérieure
Diversity and learnability in early language acquisition.
My research addresses issues of language learnability in cross-linguistic and cross-cultural settings. I follow an interdisciplinary approach, implementing computational modelling, corpus analysis and experimental methods.
Postdoctoral researcher
L’Ecole normale supérieure - PSL
Nofar.Carmeli [at] ens.fr
The fine-grained complexity of database queries.
As data analytics becomes more widespread and data becomes bigger, so grows the importance of identifying how fast we can evaluate any given query.
My research focuses on characterizing the database queries that allow a highly efficient evaluation in terms of fine-grained complexity.
Postdoctoral researcher
Université Paris Cité
sergioncbarra [at] gmail.com
MD PhD (Faculty of Medicine, University of Porto)
Sudden cardiac death. Research Program on Artificial Intelligence Applied to Cardiac Rhythm Disorders.
The identification of individuals at long-term risk of sudden cardiac death (SCD) is a difficult task. Our research group has been focusing on pre-emptive risk stratification and prevention of SCD by identifying sentinel events which frequently precede SCD, a task facilitated by recent advances in communication and remote transmission technologies. We hypothesise that it is possible to identify a high-risk group of subjects at risk of SCD and thus apply immediate preventive therapies. Application of artificial intelligence to continuous device monitoring may help identify specific patterns prior to the occurrence of a fatal event. The DAI-PP program (centralised at HEGP) has been initiated with a 2-year enrolment and 5-year follow-up of more than 5000 patients with implantable cardioverter-defibrillators and remote continuous monitoring. Furthermore, systematic collection of digital electrocardiograms is part of the prospective Paris-Sudden Death Expertise Center registry, which collects all SCD since May 2011. An improved understanding of electrical signals and their dynamics prior to the occurrence of a fatal arrhythmic event may allow a better triage and identification of patients at high-risk of events.
My portfolio includes a number of high-impact publications in the field of SCD prediction in the heart failure setting, most of which focusing on patients with clinical indication for implantable cardioverter-defibrillators or cardiac resynchronization therapy devices. Most of this work was developed while I was working at the Royal Papworth Hospital NHS Trust, in Cambridge, UK, and done in collaboration with Professors Eloi Marijon and Serge Boveda.
Postdoctoral researcher
INRIA
martinarjovsky [at] gmail.com
PhD, New York University
Out of Distribution Generalization in Machine Learning.
Machine Learning suffers from a fundamental problem: when systems are deployed in situations different than they were trained on, they can fail catastrophically. I aim to improve a large family of these problems by leveraging ideas from causality, deep learning, and the recognition of invariant statistical patterns. I am particularly focused in high-dimensional nonlinear problems (such as vision, language, and speech).
Postdoctoral researcher
L’Ecole normale supérieure - PSL
antoine.baker59 [at] gmail.com
PhD Ecole Normale Supérieure de Lyon (2011)
Compositional inference.
Approximate message passing algorithms, first applied to compressed sensing, are a promising approach to solve high-dimensional inference problems: their performance can be theoretically predicted and they often reach the optimal reconstruction error. We develop a modular package to ease the implementation of such algorithms: the user only has to declare the model then the inference, entropy estimation and prediction of performance are fully automated.