2.1.1 Application context: richer data in health and social sciences

Opportunistic data accumulations, often observational, bare great promises for social and health sciences. But the data are too big and complex for standard statistical methodologies in these sciences.

2.1.2 Related data-science challenges

7.1.1 Scikit-learn

• Keywords:
  Clustering, Classification, Regression, Machine learning
• Scientific Description:
  Scikit-learn is a Python module integrating classic machine learning algorithms in the tightly-knit scientific Python world. It aims to provide simple and efficient solutions to learning problems, accessible to everybody and reusable in various contexts: machine-learning as a versatile tool for science and engineering.
• Functional Description:

  Scikit-learn can be used as a middleware for prediction tasks. For example, many web startups adapt Scikitlearn to predict buying behavior of users, provide product recommendations, detect trends or abusive behavior (fraud, spam). Scikit-learn is used to extract the structure of complex data (text, images) and classify such data with techniques relevant to the state of the art.

  Easy to use, efficient and accessible to non datascience experts, Scikit-learn is an increasingly popular machine learning library in Python. In a data exploration step, the user can enter a few lines on an interactive (but non-graphical) interface and immediately sees the results of his request. Scikitlearn is a prediction engine . Scikit-learn is developed in open source, and available under the BSD license.

• URL:
  http://­scikit-learn.­org
• Publications:
  hal-00650905, hal-00856511, hal-01093971
• Contact:
  Gael Varoquaux
• Participants:
  Thomas Moreau, Jerome Dockes, Alexandre Gramfort, Bertrand Thirion, Gael Varoquaux, Loic Esteve, Olivier Grisel, Guillaume Lemaitre, Jeremie Du Boisberranger, Julien Jerphanion
• Partners:
  Axa, BNP Parisbas Cardif, Dataiku, Nvidia, Chanel, Probabl

7.1.2 joblib

• Keywords:
  Parallel computing, Cache
• Functional Description:
  Facilitate parallel computing and caching in Python.
• URL:
  https://­joblib.­readthedocs.­io/­en/­latest/
• Contact:
  Gael Varoquaux
• Participant:
  Thomas Moreau
• Partner:
  Probabl

7.1.3 skrub

• Keyword:
  Data analysis
• Functional Description:
  Joins, aggregates, and vectorizes tables to enable statistical learning, including with badly formated entries
• URL:
  https://­skrub-data.­org
• Contact:
  Gael Varoquaux
• Participants:
  Jerome Dockes, Riccardo Cappuzzo
• Partner:
  Probabl

Tabular deep learning

Neural networks traditionally underperform tree-based learners on tabular data. However, Holzmuller et al3 show that an array of modifications (initializations, learning-rate scheduler, feature standardization...), enables classic architectures (such as the multi-layer perceptron) to catch up. This work suggests that defaults must be adapted to the data modality, and tables call for new defaults.

Table foundation models

Much of the success of deep learning has been driven by the ability to reuse pretrained models –fitted on very large datasets, foundation models pushing this idea very far with models that provide background information useful for a wide variety of downstream tasks. A crucial part of these foundation model is the attention mechanism, stacked in a transformer architecture, that bring associative memory to the inputs by contextualizing them.

With the CARTE model 4, we adapted these ideas to tables. The strings –in the tables entries and column names– give the information that enables transfer from one table to another: data semantics. Here, key is to have an architecture that 1) models both strings and numerical values 2) applies to any set of tables while using the column names to route the information. For this purpose, CARTE uses a new dedicated attention mechanism that accounts for column names. It is pre-trained on a very large knowledge base. As a result, it outperform the best models (including tree-based models) in small sample settings (up to $n=2000$).

This result is very significant as it opens the door to foundation models for tables. It is giving birth to a very active line of research.

Prediction with missing values

Asymptotic results shows that to predict well with missing values, it is neither necessary nor sufficient to impute well these missing values by their most-likely value. Le Morvan et al5 studied the finite-sample question empirically, in the missing at random settings where, in theory, imputation in most likely to give benefits. Results show that indeed, better recovery of missing values leads to better prediction, but with diminishing returns: a large improvement in recovery quality –which typically comes at a sizable computational cost– leads to a small improvement in prediction accuracy. Additionally, the more flexible the final learner, the weaker the link is. However, adding a missing-value indicator, an extra column that indicates which values have been imputed, is always beneficial.

Assessment of large language models

Large language models (LLMs), such as chatGPT, may produce answers that are plausible but not factually correct, the so-called “hallucinations”. A variety of approach try to assess how likely a statement is to be true, for instance by sampling multiple responses from the language model. However, the challenge is to threshold these assessments, or assign a probability of correctness.

Chen et al1 investigates the confidence of LLMs in their answers. The work shows that the probabilities computed are not only overconfident, but also that there is heterogeneity (grouping loss): on some groups of queries the overconfidence is more pronounced than on others. For instance, for an answer on a notable individual, the LLMs' confidence is reasonably calibrated if the individual is from the United States, but severely overconfident for individuals from South East Asia (fig:llmconfidencenationality). Characterizing the corresponding groups opens the door to correcting the corresponding bias, a “reconfidencing” procedure.

   

Show image description

A reliability plot, giving the fraction of true statements as a function of the predicted probability for dates of birth, in replies on individuals of different nationality.

Show image description

A reliability plot, giving the fraction of true statements as a function of the predicted probability for dates of birth, in replies on individuals of different nationality.

Causal machine learning on large scale observational data

Causal approaches offer a compromise between purely predictive machine learning models that have no causal interpretation and randomized experiments that are costly and difficult to organize. Causal machine learning proposes a framework of strong assumptions to assess the causal effect of a treatment on an outcome. Those assumptions focus on the inclusion of confounding variables in the model and a non-zero probability to get the treatment for any unit in the dataset.

Dumas et al2 apply this strategy to systematically analyze the impact of chronic diseases and medications at the time of breast cancer (BC) diagnosis on cancer survival using the French Social Security data (SNDS). This would be infeasible and inefficient to do it on actual BC patients, for cost and ethical considerations, but the scale and the exhaustivity of the SNDS data enables such a study, at least to narrow down potential therapeutic candidates. In the analysis of a cohort of 235,368 French women and 288 medications with sufficient subcohort size to draw statistical conclusions, several medications have a statistically significant positive or negative effect on BC survival. Those results should not be directly interpreted as candidates for additional treatment after the BC diagnosis as chronic conditions pre-existing the diagnosis are considered here, but offer some insights about potential drug interactions or mechanism that affect the onset of BC, in particular through the immune system. This large-scale systematic study also provides a proof-of-concept of the relevance and the precision of medical knowledge that could be extracted from large claim or EHR datasets.

Reinforcement learning for adaptive recommendation of learning resources

Massive Open Online Courses (MOOCs) have greatly contributed to making education more accessible. However, many MOOCs maintain a rigid, one-size-fits-all structure that fails to address the diverse needs and backgrounds of individual learners. Learning path personalization aims to address this limitation, by tailoring sequences of educational content to optimize individual student learning outcomes. Existing approaches, however, often require either massive student interaction data or extensive expert annotation, limiting their broad application.

Vassoyan et al. 6 have framed learning path personalization as a partially observable Markov decision process. This is actually the first RL environment for dynamic cognitive diagnosis, where we assume that students learn when we show them documents within their frontier of knowledge (i.e. zone of proximal development) and our goal is to optimize their learning outcomes. By propagating information on a bipartite graph of keywords and documents, we could learn a policy (using REINFORCE algorithm) for selecting the best learning resource for learning a topic. By using word embeddings on the content of documents, we could alleviate item cold-start. We conducted experiments with simulated students on a real corpus of MOOCs. We went deeper in reducing the data needed to provide relevant recommendations of documents: dozens of episodes instead of thousands of episodes. We also showed that our method can generalize to unseen corpuses of documents. This is a collaboration with Anan Schütt & Elisabeth André from U. Augsburg, Arun Narayanan from U. Pittsburgh, and Nicolas Vayatis from Centre Borelli.

Releases of scikit-learn

With 3 major releases in 2024 (1.4 in Jan, 1.5 in May, and 1.6 in December), Scikit-learn is always improving, adding features for better and easier machine learning in Python. We list below a few highlights that are certainly not exhaustive but illustrate the continuous progress made.

Show image description

Monotonic constraints in trees clip out the bumps when fitting the addition of a sine wave and a constantly increasing slope

Show image description

Monotonic constraints in trees clip out the bumps when fitting the addition of a sine wave and a constantly increasing slope

• Controling classifier threshold
  Given a fitted classifier, the FinedThresholdClassifier and TunedThresholdClassifier can adjust the threshold to maximize a given utility, either set theoretically with a cost matrix, or empirically to minimize the cost on a validation set.
• FrozenEstimator
  An already-fitted estimator can be given to FrozenEstimator so to have an object that is no longer modified at fit time. This is useful to inject in pipelines pre-trained models as it enables reusing standard model evaluation tools.
• Categorical support
  If an input is given as a dataframe (pandas or polars) with some columns typed as categories, HistGradientBoosting classifier and regressor will use a categorical splitter in the trees for these.
• Polars output
  using the set_output method of an estimator, transformers can output a polars dataframe, respecting the column names of the input if any.
• Missing value support
  Random Forest and Extra Trees now support missing-values natively, fitting them by special-casing them in the construction of the tree (a strategy sometimes called “Missing Incorporated Attribute").
• Monotonic constraints in tree models
  Different tree-based models (HistGradientBoosting regressor and classifier, random forests, extra trees) can now take constraints to force the prediction function to be monotonous along given features (fig:monotonictres).

skrub

The first release of skrub was late 2023. There have been 3 releases in 2024, leading to 0.4 in December 2024. Skrub is a package to facilitate machine learning on tables. The major features added in 2024 are:

• TableReport
  The TableReport gives an interactive display of dataframes, enabling inspection of the different columns and their distributions, that can be easily embedded, including in the programming environment used by data scientists.
• TextEncoder
  The TextEncoder uses a pretrained deep-learning language model to embed strings in a given column.
• tabular_learner
  The tabular_learner function builds a preprocessing pipeline that encodes messy data frames in a way that is well suited for a given a predictor.

joblib

joblib is a very simple computation engine in Python that is massively used worldwide, including as a dependency of packages such as scikit-learn for parallel computing.

Release 1.4 (May 2024). Many changes to follow evolutions of the ecosystem and improve behaviors (eg better error handling). Major changes are:

• Allowing to cache coroutines
• Optional unordered execution of parallel loop, to better use multiple CPUs

Probabl

Probabl is an Inria spin-off in which Gaël Varoquaux has 10% of his time allocated. Probabl's mission is to develop and make sustainable an ecosystem of data-science commons. Probabl is the larger employer of scikit-learn maintainers. It builds a commercial offer around the scikit-learn ecosystem for the entreprise. Gaël Varoquaux is the point of contact at Soda.

Pass Culture

Within the Ministry of Culture-Inria convention, Samuel Girard and Jill-Jênn Vie have been involved in a partnership with Pass Culture (used by 3M students in France) to improve the diversity of their recommendations (12 months, started in June 2024). We hired an engineer, Hiba Bederina, from June 2024.

Collaboration with Ministère de la Santé

We have a 2-year long collaboration with Ministère de la Santé (HAS) on using the national healthcare data for prevention and policy evaluation. Gaël Varoquaux and Judith Abecassis are in charge at Soda.

Collaboration with public interest group Pix

Jill-Jênn Vie got a Paris Region PhD 2023 funding with Pix (certification of digital competencies, 6M active users), about optimizing human learning using reinforcement learning. Samuel Girard 's PhD is currently on this funding (105k from région Île-de-France, 20k from Pix).

Plan de relance with Dataiku

Soda had a 24 months post-doc funded by “plan de relance” jointly with Dataiku on using embeddings for database analytics. The post-doc started beginning of November 2022 and ended in October 2024. Gaël Varoquaux is in charge at Soda.

10.1.1 Inria associate team not involved in an IIL or an international program

10.2.1 Horizon Europe

PEPR Santé Numérique

Soda is part of the “PEPR Santé Numérique” in the SMATCH subgroup that focuses on evidence of clinical efficacy. Soda will address two questions. The first question, addressed in collaboration with the PreMedical team, is that of external validity of randomized trials: how much is the treatment effect measured in a randomized clinical trial affected by the sampling bias of the trial, the difference between the study population and the intended target population. The second question, addressed in collaboration with the Heka team, is that of defining guidelines to evaluate software as a medical device. One particular challenge that we will tackle is to give procedures and recommendations to evaluate an update to a software used in clinical decision making using historical data rather than a trial. The project started end of 2023. Gaël Varoquaux is in charge at Soda, and Judith Abecassis is also supervising.

Project Partages

“Partages” is a large project funded by BPI France to develop digital commons for medical text analysis. In particular, the project will create material suitable for fine-tuning or aligning language models to perform best on French medical texts. Beyond the medical terms, there are specific challenges of clinical texts: these often result from scanning notes that have been taken fast, full of context-specific abbreviations and typos. The role of Soda is to design data-augmentation routine that help making language models robust to these challenges. The project started end of 2024. Gaël Varoquaux is in charge at Soda, and Judith Abecassis is also supervising.

French National IA commission

Gaël Varoquaux was an expert at the French National commission that lasted from September 2023 to March 2024. The commission auditioned hundreds of experts and did bibliographic research to advise on all aspects of AI and society, having been tasked by the government to advise on public policy around AI. In March, the commission handed in a general report, with 150 pages of high-level analysis, as well as technical recommendations for the French administration.

International Scientific Report on the Safety of Advanced AI

Gaël Varoquaux is an expert for the UK AI Safety Institute and for the International Scientific Report on the Safety of Advanced AI 36. A panel of international experts from all countries are working together with a team of experts to consolidate analysis around AI safety. The report touches upon all aspects of risks of AI (privacy, fairness, cybersecurity, bio weapons, loss of control...) as well as the current scientific evidence on their evolutions and possible mitigations. It is meant to ground public policy across the world on scientific evidence. An interim report is available 36, but the final report is still in progress.

11.1.1 Scientific events: organisation

11.1.2 Scientific events: selection

11.1.3 Journal

11.1.4 Invited talks

• Gaël Varoquaux

  Keynotes: TRL workshop NeurIPS, ODSC Europe, Epiclin, Dataiku Technical Kick Off, Morocco AI, MIDL, ICCR, P16 Kick Off

  Other invited talks dotAI, Health Data Hub scientific days, SESSTIM seminar, ENS Lyon physics seminar, Mila biomedical group

• Judith Abécassis
  Young Statisticians and Probabilists (YSP) Days, Sacl-AI for Science Workshop, workshop "Causal inference in Genetics", les Treilles.
• Marine Le Morvan
  MIA Paris-Saclay seminar (INRAE), Statistics and Computer Science Day (IHES)
• Jill-Jênn Vie
  Conseil scientifique de l'Éducation nationale, AI in education workshop (Inria Bordeaux)

11.1.5 Scientific expertise

• Gaël Varoquaux
  Comité Scientifique ANR Techniques Spécifiques de l'IA, Comité Scientifique DATAE
• Judith Abécassis
  Scientific expert for the call "AGIR EN SANTÉ PUBLIQUE (AGIR-SP)" INCa
• Jill-Jênn Vie
  Organisation internationale de la francophonie, conseil scientifique MonProjetSup

11.1.6 Research administration

• Gaël Varoquaux
  Scientific president of the ClusterAI submission for Université Paris Saclay

Courses

• Gaël Varoquaux
   

  • AI on tabular data, Ellis Doctoral Symposium, 1.5h
  • Preparing tabular data for machine learning, tutorial, AutoML conference, 1.5h
  • Learning on messy tabular data, Hi-Paris Summer School, 1.5h
  • Machine learning, Inria academy, executives from the French ministry of defense 1h
• Marine Le Morvan
   

  • Deep Learning, Ecole Polytechnique, 27h
  • Refresher Course in Artificial Intelligence, Ecole Polytechnique, 15h
  • Statistics and machine learning with missing values, Université Paris Dauphine, 6h
• Jill-Jênn Vie
   

  • INF471S ICPC–SWERC training (advanced algorithms), École polytechnique, 60h
  • Préparation au SWERC, ENS Paris-Saclay, 21h eq. TD
• Judith Abécassis
   

  • Causal Inference DS-UA 9201, NYU Paris, Spring 2024, 56h eq. TD
  • Causal Inference DS-UA 9201 (with Houssam Zenati, MIND team), NYU Paris, Fall 2024, 48h eq. TD

E-learning

• Machine learning with Scikit-learn MOOC
  40 hours of learning starting as an introduction to machine learning and covering more advanced topics such as data preparation and model selection. Accessible on inria.github.io/scikit-learn-mooc, and designed by Loïc Esteve , Arturo Amor , Guillaume Lemaître , Olivier Grisel , Gaël Varoquaux .

11.2.1 Supervision

• Gaël Varoquaux
  PhD advisor for Celestin Eve (50%), Sébastien Melo (60%), Meilame Tayebjee (25%) Julie Alberge (30%), Jovan Stojanovic (50%), Félix Lefebvre (100%), Alexandre Perez (50%), Léo Grinsztajn (50%)
• Judith Abécassis
  PhD advisor for Julie Alberge (70%)
• Marine Le Morvan
  PhD advisor for Alexandre Perez (50%), Sébastien Melo (40%)
• Jill-Jênn Vie
  PhD advisor for Jean Vassoyan (33%), Marie Generali (33%), Samuel Girard (33%)

11.2.2 Juries

• Gaël Varoquaux
  PhD committee for Hugo Thimonier (éxaminateur)
• Marine Le Morvan
   

  • PhD comittee for Gabriel Damay (éxaminateur)
  • jury pour concours CRCN/ISFP Saclay
  • comité de sélection poste professeur assistant en apprentissage statistique à Polytechnique

11.3.1 Productions (articles, videos, podcasts, serious games, ...)

• Gaël Varoquaux

  Scientific chronicles in Les Échos

  Podcasts IA, pas que de la data; Data driven 101; Dialogue Machine

• Judith Abécassis
  article dans le journal TELECOM "Quels usages pour l’intelligence artificielle en oncologie ? de la recherche au patient"

11.3.2 Participation in Live events

• Gaël Varoquaux
   

  • Panel on AI at BPI's BIG forum, one of the largest forum on investments in Tech in France
  • Panel on AI at the Paris “Chambre de Commerce et de l'Industrie”
  • Panel on AI and creative and cultural industries, French embassy Berlin
  • Panel on sustainability and AI, AI pulse
• Judith Abécassis
   

  • Round table "IA et parcours de soins", espace PariSanté Campus, salon SantExpo
  • conference "IA et Santé : Des modèles prédictifs aux modèles prescriptifs, interprétabilité et explicabilité", Startup accelerator for health prevention, BPI France, PariSante Campus (via Inria Academy)
  • webinar "IA et Santé : Des modèles prédictifs aux modèles prescriptifs" Ramsay Santé, Inria Academy
• Jill-Jênn Vie
   

  • Panel at UNESCO: “AI4T: AI for and by teachers & Pix: assessing digital skills in a changing world”, Digital competencies for teachers and school students seminar
  • Atelier “Intelligence artificielle et éducation”, Colloque In Fine (plan de formation à destination des personnels de l'éducation nationale), Futuroscope
  • Coach of École polytechnique at International Competitive Programming Contest (ICPC): 2 Silver medals (3rd & 4th places) at Southwestern Regionals 2024 in January 2024, admitted to ICPC World Finals in Astana, Kazakhstan in September 2024; won Southwestern Regionals 2025 in December 2024 (1st, 3rd, and 21st places), Gold and Silver medal.
  • Panel on AI and human learning, Numérique en communs, November 2024

11.3.3 Others science outreach relevant activities

• Jill-Jênn Vie
  Girls Can Code! 2-day girls-only autumn school for discovering programming in November 2024 in École polytechnique. 50 participants from middle school and high school. Organized with non-profit organization Prologin.

International journals

• 8 articleB.Bénédicte Colnet, J.Julie Josse, G.Gaël Varoquaux and E.Erwan Scornet. Reweighting the RCT for generalization: finite sample error and variable selection.Journal of the Royal Statistical Society: Series A Statistics in SocietyMay 2024HALDOI
• 9 articleB.Bénédicte Colnet, I.Imke Mayer, G.Guanhua Chen, A.Awa Dieng, R.Ruohong Li, G.Gaël Varoquaux, J.-P.Jean-Philippe Vert, J.Julie Josse and S.Shu Yang. Causal inference methods for combining randomized trials and observational studies: a review.Statistical Science2024. In press. HAL
• 10 articleE.Elise Dumas, B.Beatriz Grandal Rejo, P.Paul Gougis, S.Sophie Houzard, J.Judith Abécassis, F.Floriane Jochum, B.Benjamin Marande, A.Annabelle Ballesta, E.Elaine del Nery, T.Thierry Dubois, S.Samar Alsafadi, B.Bernard Asselain, A.Aurélien Latouche, M.Marc Espie, E.Enora Laas, F.Florence Coussy, C.Clémentine Bouchez, J.-Y.Jean-Yves Pierga, C.Christine Le Bihan-Benjamin, P.-J.Philippe-Jean Bousquet, J.Judicaël Hotton, C.-A.Chloé-Agathe Azencott, F.Fabien Reyal and A.-S.Anne-Sophie Hamy. Concomitant medication, comorbidity and survival in patients with breast cancer.Nature Communications151April 2024, 2966HALDOI
• 11 articleT.T. Jolivet, J.-B.J-B. Julla, Y.Y. Abouleka, A.A. Berges, G.G. Varoquaux, J.J. Abecassis, J.J. Alberge, T.T. Petit Jean, E.E. Larger, A.A. Hartemann, J.-F.J-F. Gautier, L.L. Potier, C.C. Estellat and F.F. Tubach. Taux de mortalité et facteurs de risque associés chez les patients diabétiques hospitalisés pour plaie du pied : étude de cohorte historique sur l'EDS-APHP.Journal of Epidemiology and Population Health72March 2024, 202255HALDOI
• 12 articleJ.Julie Josse, J. M.Jacob M. Chen, N.Nicolas Prost, G.Gaël Varoquaux and E.Erwan Scornet. On the consistency of supervised learning with missing values.Statistical Papers659March 2024, 5447-5479HALDOI
• 13 articleL.Lena Maier-Hein, A.Annika Reinke, P.Patrick Godau, M.Minu Tizabi, F.Florian Buettner, E.Evangelia Christodoulou, B.Ben Glocker, F.Fabian Isensee, J.Jens Kleesiek, M.Michal Kozubek, M.Mauricio Reyes, M.Michael Riegler, M.Manuel Wiesenfarth, A. E.A Emre Kavur, C.Carole Sudre, M.Michael Baumgartner, M.Matthias Eisenmann, D.Doreen Heckmann-Nötzel, T.Tim Rädsch, L.Laura Acion, M.Michela Antonelli, T.Tal Arbel, S.Spyridon Bakas, A.Arriel Benis, M.Matthew Blaschko, M. J.M Jorge Cardoso, V.Veronika Cheplygina, B.Beth Cimini, G.Gary Collins, K.Keyvan Farahani, L.Luciana Ferrer, A.Adrian Galdran, B.Bram van Ginneken, R.Robert Haase, D.Daniel Hashimoto, M.Michael Hoffman, M.Merel Huisman, P.Pierre Jannin, C.Charles Kahn, D.Dagmar Kainmueller, B.Bernhard Kainz, A.Alexandros Karargyris, A.Alan Karthikesalingam, F.Florian Kofler, A.Annette Kopp-Schneider, A.Anna Kreshuk, T.Tahsin Kurc, B.Bennett Landman, G.Geert Litjens, A.Amin Madani, K.Klaus Maier-Hein, A.Anne Martel, P.Peter Mattson, E.Erik Meijering, B.Bjoern Menze, K.Karel Moons, H.Henning Müller, B.Brennan Nichyporuk, F.Felix Nickel, J.Jens Petersen, N.Nasir Rajpoot, N.Nicola Rieke, J.Julio Saez-Rodriguez, C.Clara Sánchez, S.Shravya Shetty, M.Maarten van Smeden, R.Ronald Summers, A.Abdel Taha, A.Aleksei Tiulpin, S.Sotirios Tsaftaris, B.Ben van Calster, G.Gaël Varoquaux and P.Paul Jäger. Metrics reloaded: recommendations for image analysis validation.Nature Methods212February 2024, 195-212HALDOI
• 14 articleS.Sunghong Park, M. J.Myung Jun Kim, K.Kanghee Park and H.Hyunjung Shin. Mutual Domain Adaptation.Pattern Recognition145January 2024, 109919HALDOI
• 15 articleA. L.Ana Luísa Pinho, H.Hugo Richard, A. F.Ana Fernanda Ponce, M.Michael Eickenberg, A.Alexis Amadon, E.Elvis Dohmatob, I.Isabelle Denghien, J. J.Juan Jesús Torre, S.Swetha Shankar, H.Himanshu Aggarwal, A.Alexis Thual, T.Thomas Chapalain, C.Chantal Ginisty, S.Séverine Becuwe-Desmidt, S.Séverine Roger, Y.Yann Lecomte, V.Valérie Berland, L.Laurence Laurier, V.Véronique Joly-Testault, G.Gaëlle Médiouni-Cloarec, C.Christine Doublé, B.Bernadette Martins, G.Gaël Varoquaux, S.Stanislas Dehaene, L.Lucie Hertz-Pannier and B.Bertrand Thirion. Individual Brain Charting dataset extension, third release for movie watching and retinotopy data.Scientific Data 11590June 2024HAL
• 16 articleR.Russell Poldrack, C.Christopher Markiewicz, S.Stefan Appelhoff, Y.Yoni Ashar, T.Tibor Auer, S.Sylvain Baillet, S.Shashank Bansal, L.Leandro Beltrachini, C.Christian Benar, G.Giacomo Bertazzoli, S.Suyash Bhogawar, R.Ross Blair, M.Marta Bortoletto, M.Mathieu Boudreau, T.Teon Brooks, V.Vince Calhoun, F. M.Filippo Maria Castelli, P.Patricia Clement, A.Alexander Cohen, J.Julien Cohen-Adad, S.Sasha d'Ambrosio, G.Gilles de Hollander, M.María de la Iglesia-Vayá, A.Alejandro de la Vega, A.Arnaud Delorme, O.Orrin Devinsky, D.Dejan Draschkow, E. P.Eugene Paul Duff, E.Elizabeth Dupre, E.Eric Earl, O.Oscar Esteban, F.Franklin Feingold, G.Guillaume Flandin, A.Anthony Galassi, G.Giuseppe Gallitto, M.Melanie Ganz, R.Rémi Gau, J.James Gholam, S.Satrajit Ghosh, A.Alessio Giacomel, A.Ashley Gillman, P.Padraig Gleeson, A.Alexandre Gramfort, S.Samuel Guay, G.Giacomo Guidali, Y.Yaroslav Halchenko, D.Daniel Handwerker, N.Nell Hardcastle, P.Peer Herholz, D.Dora Hermes, C.Christopher Honey, R.Robert Innis, H.-I.Horea-Ioan Ioanas, A.Andrew Jahn, A.Agah Karakuzu, D.David Keator, G.Gregory Kiar, B.Balint Kincses, A.Angela Laird, J.Jonathan Lau, A.Alberto Lazari, J. H.Jon Haitz Legarreta, A.Adam Li, X.Xiangrui Li, B.Bradley Love, H.Hanzhang Lu, E.Eleonora Marcantoni, C.Camille Maumet, G.Giacomo Mazzamuto, S.Steven Meisler, M.Mark Mikkelsen, H.Henk Mutsaerts, T.Thomas Nichols, A.Aki Nikolaidis, G.Gustav Nilsonne, G.Guiomar Niso, M.Martin Norgaard, T.Thomas Okell, R.Robert Oostenveld, E.Eduard Ort, P.Patrick Park, M.Mateusz Pawlik, C.Cyril Pernet, F.Franco Pestilli, J.Jan Petr, C.Christophe Phillips, J.-B.Jean-Baptiste Poline, L.Luca Pollonini, P. R.Pradeep Reddy Raamana, P.Petra Ritter, G.Gaia Rizzo, K.Kay Robbins, A.Alexander Rockhill, C.Christine Rogers, A.Ariel Rokem, C.Chris Rorden, A. M.Alexandre M Routier, J. M.Jose Manuel Saborit-Torres, T.Taylor Salo, M.Michael Schirner, R.Robert Smith, T.Tamas Spisak, J.Julia Sprenger, N.Nicole Swann, M.Martin Szinte, S.Sylvain Takerkart, B.Bertrand Thirion, A.Adam Thomas, S.Sajjad Torabian, G.Gaël Varoquaux, B.Bradley Voytek, J.Julius Welzel, M.Martin Wilson, T.Tal Yarkoni and K.Krzysztof Gorgolewski. The Past, Present, and Future of the Brain Imaging Data Structure (BIDS).Imaging Neuroscience22024, 1-19HALDOI
• 17 articleJ.Jingang Qu, S.Soleiman Yousef, T.Thibault Faney, J.-C.Jean-Charles de Hemptinne and P.Patrick Gallinari. NNEoS : Neural network-based thermodynamically consistent equation of state for fast and accurate flash calculations.Applied Energy374November 2024, 124025HALDOI
• 18 articleC.Clémence Réda, J.-J.Jill-Jênn Vie and O.Olaf Wolkenhauer. Comprehensive evaluation of pure and hybrid collaborative filtering in drug repurposing.Scientific Reports2025. In press. HAL
• 19 articleC.Clémence Réda, J.-J.Jill-Jênn Vie and O.Olaf Wolkenhauer. Joint Embedding-Classifier Learning for Interpretable Collaborative Filtering.BMC Bioinformatics2024. In press. HAL
• 20 articleC.Clémence Réda, J.-J.Jill-Jênn Vie and O.Olaf Wolkenhauer. stanscofi and benchscofi: a new standard for drug repurposing by collaborative filtering.Journal of Open Source Software993January 2024, 5973HALDOI
• 21 articleA.Annika Reinke, M. D.Minu D. Tizabi, M.Michael Baumgartner, M.Matthias Eisenmann, D.Doreen Heckmann-Nötzel, A. E.A. Emre Kavur, T.Tim Rädsch, C. H.Carole H. Sudre, L.Laura Acion, M.Michela Antonelli, T.Tal Arbel, S.Spyridon Bakas, A.Arriel Benis, M.Matthew Blaschko, F.Florian Buettner, M. J.M. Jorge Cardoso, V.Veronika Cheplygina, J.Jianxu Chen, E.Evangelia Christodoulou, B. A.Beth A. Cimini, G. S.Gary S. Collins, K.Keyvan Farahani, L.Luciana Ferrer, A.Adrian Galdran, B.Bram van Ginneken, B.Ben Glocker, P.Patrick Godau, R.Robert Haase, D. A.Daniel A. Hashimoto, M. M.Michael M. Hoffman, M.Merel Huisman, F.Fabian Isensee, P.Pierre Jannin, C. E.Charles E. Kahn, D.Dagmar Kainmueller, B.Bernhard Kainz, A.Alexandros Karargyris, A.Alan Karthikesalingam, H.Hannes Kenngott, J.Jens Kleesiek, F.Florian Kofler, T.Thijs Kooi, A.Annette Kopp-Schneider, M.Michal Kozubek, A.Anna Kreshuk, T.Tahsin Kurc, B. A.Bennett A. Landman, G.Geert Litjens, A.Amin Madani, K.Klaus Maier-Hein, A. L.Anne L. Martel, P.Peter Mattson, E.Erik Meijering, B.Bjoern Menze, K. G.Karel G. M. Moons, H.Henning Müller, B.Brennan Nichyporuk, F.Felix Nickel, J.Jens Petersen, S. M.Susanne M. Rafelski, N.Nasir Rajpoot, M.Mauricio Reyes, M. A.Michael A. Riegler, N.Nicola Rieke, J.Julio Saez-Rodriguez, C. I.Clara I. Sánchez, S.Shravya Shetty, R. M.Ronald M. Summers, A. A.Abdel A. Taha, A.Aleksei Tiulpin, S. A.Sotirios A. Tsaftaris, B.Ben van Calster, G.Gaël Varoquaux, Z. R.Ziv R. Yaniv, P. F.Paul F. Jäger and L.Lena Maier-Hein. Understanding metric-related pitfalls in image analysis validation.Nature Methods212024, 182–194HALDOI
• 22 articleV.Viktor Zaverkin, D.David Holzmüller, H.Henrik Christiansen, F.Federico Errica, F.Francesco Alesiani, M.Makoto Takamoto, M.Mathias Niepert and J.Johannes Kästner. Uncertainty-biased molecular dynamics for learning uniformly accurate interatomic potentials.npj Computational Materials101April 2024, 83HALDOI

International peer-reviewed conferences

• 23 inproceedingsL.Lihu Chen, G.Gaël Varoquaux and F. M.Fabian M. Suchanek. Learning High-Quality and General-Purpose Phrase Representations.EACL 2024 - The 18th Conference of the European Chapter of the Association for Computational LinguisticsLa Valette, MaltaMarch 2024HAL
• 24 inproceedingsS.Samuel Girard, J.-J.Jill-Jênn Vie, F.Françoise Tort and A.Amel Bouzeghoub. Optimizing human learning using reinforcement learning.Educational Data Mining 2024Atlanta (USA), United StatesJuly 2024HAL
• 25 inproceedingsE. N.Erva Nihan Kandemir, J.-J.Jill-Jênn Vie, A. H.Adam Hegel Sanchez Ayte, O.Olivier Palombi and F.Franck Ramus. Adaptation of the Multi-Concept Multivariate Elo Rating System to Medical Students' Training Data.Proceedings of the 14th Learning Analytics and Knowledge ConferenceLAK 2024 - The 14th Learning Analytics and Knowledge ConferenceKyoto, Japan2024, 1-10HALDOI
• 26 inproceedingsM. J.Myung Jun Kim, L.Léo Grinsztajn and G.Gaël Varoquaux. CARTE: Pretraining and Transfer for Tabular Learning.Proceedings of Machine Learning ResearchForty-first International Conference on Machine Learning, ICML 2024235Vienna, AustriaJuly 2024HAL
• 27 inproceedingsJ.-J.Jill-Jênn Vie, Y.Yong Zheng, F.Fabrice Popineau, Y.Yizhu Gao, S.Samuel Girard and H.Hisashi Kashima. Optimizing Human Learning. 4th Workshop eliciting Adaptive Sequences for Learning and Educational RecSys (WASL 2024).Learning Analytics and Knowledge 2024Kyoto, JapanMarch 2024HAL

Conferences without proceedings

• 28 inproceedingsJ.Judith Abécassis, T.Théo Jolivet, A.Audrey Bergès, E.Elise Liu, J.-B.Jean-Baptiste Julla, Y.Yawa Abouleka, J.Julie Alberge, I.Isabel Bonnetier, T.Thomas Petit-Jean, R.Romain Bey, C.Candice Estellat, F.Florence Tubach, G.Gaël Varoquaux and L.Louis Potier. Operational challenges of building a million-patient cohort from EHRs: The COhort of DIabetic patients (CODIA) on the AP-HP EDS.journée de l'Atelier TIDS (Traitement Informatique des Données de Santé) du GdR MaDICSParis (PariSanté Campus), FranceOctober 2024HAL
• 29 inproceedingsL.Lihu Chen, A.Alexandre Perez-Lebel, F.Fabian Suchanek and G.Gaël Varoquaux. Reconfidencing LLM Uncertainty from the Grouping Loss Perspective.EMNLP 2024 - Conference on Empirical Methods in Natural Language ProcessingMiami, United StatesarXiv2024HALDOI
• 30 inproceedingsD.David Holzmüller, L.Léo Grinsztajn and I.Ingo Steinwart. Better by Default: Strong Pre-Tuned MLPs and Boosted Trees on Tabular Data.Neural Information Processing SystemsVancouver (BC), Canada2024HALDOI
• 31 inproceedingsJ.Jean Vassoyan, A.Anan Schütt, J.-J.Jill-Jênn Vie, A.-B.Arun-Balajiee Lekshmi-Narayanan, E.Elisabeth André and N.Nicolas Vayatis. A Pre-Trained Graph-Based Model for Adaptive Sequencing of Educational Documents.NeurIPS 2024 Workshop FM-EduAssess - The First Workshop p, Large Foundation Models for Educational AssessmentVancouver, CanadaDecember 2024HAL

Scientific book chapters

• 32 inbookI.Irene Balelli, S.Safaa Al-Ali, E.Elise Dumas and J.Judith Abécassis. Causality: fundamental principles and tools.Trustworthy AI in Medical ImagingChapitre 14November 2024, 297-314HAL

Reports & preprints

• 33 miscJ.Judith Abécassis, E.Elise Dumas, J.Julie Alberge and G.Gaël Varoquaux. From prediction to prescription: Machine learning and Causal Inference.November 2024HAL
• 34 reportP.Philippe Aghion, A.Anne Bouverot, A.Arno Amabile, C.Cyprien Canivenc, G.Gilles Babinet, J.Joëlle Barral, A.Alexandra Bensamoun, N.Nozha Boujemaa, B.Bertrand Charlès, L.Luc Julia, Y.Yann Lecun, A.Arthur Mensch, C.Cédric O, I.Isabelle Ryl, F.Franca Salis-Madinier, M.Martin Tisné, G.Gaël Varoquaux, M.Marc Auberger, S.Simon Bunel, P.Philippe Chantepie, E.Eloy Dorado, E.-P.Emilie-Pauline Gallié, P.Paul Jolie, A.Arnaud Mazier, V.Vincent Montreuil, E.Erwan Paitel, T.Timothée Paris, C.Christophe Ravier, U.Ulrich Tan and L. C.Louis Charles Viossat. IA : Notre Ambition Pour La France: Commission De L'intelligence Artificielle.Gouvernement FrançaisMarch 2024HALback to text
• 35 miscJ.Julie Alberge, V.Vincent Maladière, O.Olivier Grisel, J.Judith Abécassis and G.Gaël Varoquaux. Survival Models: Proper Scoring Rule and Stochastic Optimization with Competing Risks.May 2024HALback to text
• 36 reportY.Yohsua Bengio, S.Soren Mindermann, D.Daniel Privitera, T.Tamay Besiroglu, R.Rishi Bommasani, S.Stephen Casper, Y.Yejin Choi, D.Danielle Goldfarb, H.Hoda Heidari, L.Leila Khalatbari, S.Shayne Longpre, V.Vasilios Mavroudis, M.Mazeika Mantas, K. Y.Kwan Yee Ng, C. T.Chinasa T. Okolo, D.Deborah Raji, T.Theodora Skeadas, F.Florian Tramèr, B.Bronwyn Fox, A. C.Andre Carlos Ponce de Leon Ferreira de Carvalho, M.Mona Nemer, R.Raquel Pezoa Rivera, Y.Yi Zeng, J.Juha Heikkilä, G.Guillaume Avrin, A.Antonio Krüger, B.Balaraman Ravindran, H.Hammam Riza, C.Ciarán Seoighe, Z.Ziv Katvir, A.Andrea Monti, H.Hiroaki Kitano, M.Mary Kerema, J.José Ramón, H.Haroon Sheikh, G.Gill Jolly, O.Olubunmi Ajala, D.Dominic Ligot, K. M.Kyoung Mu Lee, A.Ahmet Halit, C.Crystal Rugege, F.Fahad Albalawi, D.Denise Wong, N.Nuria Oliver, C.Christian Busch, O.Oleksii Molchanovskyi, M.Marwan Alserkal, S.Saif Khan, A.Angela Mclean, A.Amandeep Gill, B.Bayo Adekanmbi, P.Paul Christiano, D.David Dalrymple, T. G.Thomas G. Dietterich, E.Edward Felten, P.Pascale Fung, P.-O.Pierre-Olivier Gourinchas, N.Nick Jennings, A.Andreas Krause, P.Percy Liang, T.Teresa Ludermir, V.Vidushi Marda, H.Helen Margetts, J. A.John A. Mcdermid, A.Arvind Narayanan, A.Alondra Nelson, A.Alice Oh, G.Gopal Ramchurn, S.Stuart Russell, M.Marietje Schaake, D.Dawn Song, A.Alvaro Soto, L.Lee Tiedrich, G.Gaël Varoquaux, A.Andrew Yao and Y.Yaqin Zhang. International Scientific Report on the Safety of Advanced AI: interim report.Department for Science, Innovation and TechnologyMay 2024HALback to textback to textback to text
• 37 miscB.Ben van Calster, G. S.Gary S. Collins, A. J.Andrew J. Vickers, L.Laure Wynants, K. F.Kathleen F. Kerr, L.Lasai Barreñada, G.Gaël Varoquaux, K.Karandeep Singh, K. G.Karel G. M. Moons, T.Tina Hernandez-Boussard, D.Dirk Timmerman, D. J.David J. Mclernon, M.Maarten van Smeden and E. W.Ewout W. Steyerberg. Performance evaluation of predictive AI models to support medical decisions: Overview and guidance.December 2024HAL
• 38 miscR.Riccardo Cappuzzo, G.Gaël Varoquaux, A.Aimee Coelho and P.Paolo Papotti. Retrieve, Merge, Predict: Augmenting Tables with Data Lakes (Experiment, Analysis & Benchmark Paper).March 2024HAL
• 39 miscL.Lihu Chen and G.Gaël Varoquaux. What is the Role of Small Models in the LLM Era: A Survey.2024HALDOI
• 40 miscM.Marine Le Morvan and G.Gaël Varoquaux. Imputation for prediction: beware of diminishing returns.July 2024HAL
• 41 miscD.Daniel Musekamp, M.Marimuthu Kalimuthu, D.David Holzmüller, M.Makoto Takamoto and M.Mathias Niepert. Active Learning for Neural PDE Solvers.2024HALDOI
• 42 miscG.Gaël Varoquaux, A. S.Alexandra Sasha Luccioni and M.Meredith Whittaker. Hype, Sustainability, and the Price of the Bigger-is-Better Paradigm in AI.September 2024HALback to text
• 43 miscJ.Julien Zhou, P.Pierre Gaillard, T.Thibaud Rahier, H.Houssam Zenati and J.Julyan Arbel. Towards Efficient and Optimal Covariance-Adaptive Algorithms for Combinatorial Semi-Bandits.2024HAL