Gaël Varoquaux
Research & code: Research director @inria
►Data, Health, & Computer science
►Python coder, (co)founder of scikit-learn, joblib, & @probabl.bsky.social
►Sometimes does art photography
►Physics PhD
- I increasingly worry that tech is building the ideal 1984-style propaganda + surveillance machine The risk is the collapse of democracy, as the elite concentrates too much control power, making it counter-powers impossible. Democracy needs counter-powers, including activists.
- Reposted by Gaël Varoquaux
- Reposted by Gaël VaroquauxAnd here you'll find some examples on how to use expressions in practice: skrub-data.org/dev/auto_exa...
- Reposted by Gaël VaroquauxIf you're eager to try out Skrub Expressions, you can clone the dev branch of our repository: skrub-data.org/dev/
- Reposted by Gaël Varoquaux👀 This week's post is a sneak peek into the next major Skrub feature, Skrub expressions 🚀 As this is a preview of an upcoming feature, we are looking for your thoughts and feedback before release.
- Reposted by Gaël VaroquauxInria is opening a Junior Professor Position (tenure track) at Saclay on the topic of Cognitive Computational Neuroscience. This junior professorship is part of the development of computational cognitive neuroscience within @univparissaclay.bsky.social #Neurospin @neuropsi.bsky.social and #inria
- Some photography: Streets of Naples, dark and narrow flickr.com/photos/gaelv... Walking down "ScappaNapoli", a long and narrow street that cuts straight through Naples. In the distance, the sun on streets going up to Sant Elmo castle. But this street is dark and crowded. #Napoli #photograghy
- Reposted by Gaël VaroquauxOr read the documentation: skrub-data.org/stable/refer...
- Reposted by Gaël VaroquauxCheck out a demo and more examples: skrub-data.org/skrub-report...
- Reposted by Gaël VaroquauxThe Skrub TableReport is a lightweight tool that allows to get a rich overview of a table quickly and easily. ✅ Filter columns 🔎 Look at each column's distribution 📊 Get a high level view of the distributions through stats and plots, including correlated columns 🌐 Export the report as html
- #ICLR2025 Marine Le Morvan presents "Imputation for prediction: beware of diminishing returns": poster Thu 24th arxiv.org/abs/2407.19804 Concludes 6 years of research on prediction with missing values: Imputation is useful but improvements are expensive, while better learners yield easier gains.
- Reposted by Gaël VaroquauxAnd if you're not familiar with what Skrub is all about, you might want to check out our introductory slide deck here: skrub-data.org/skrub-materi...
- Reposted by Gaël Varoquaux🚀 The Skrub learning materials website is now live at: skrub-data.org/skrub-materi... Here you'll find introductory talks and tutorials about Skrub, along with notebooks and blog posts showcasing the features of the library. Bookmark it to not miss any update 👀
- Reposted by Gaël Varoquaux🚨 1 day left ! Register as a speaker for PyData Paris 🚨 pydata.org/paris2025/cfp
- Reposted by Gaël Varoquaux[ #VeilleESR #LRU ] Les contrats d’objectifs, de moyens et de performance (COMP) conclus entre l’État et les établissements d'ESR - @ccomptes.fr TLDR : Fabriqués dans le secret et dans l'urgence, les COMP n'ont aucune chance de marcher. C'est pourquoi ils doivent être généralisés.
- Amis chercheurs Français, A partir de quand les arguments sur la propriété intellectuelle dans certains EPST rentrent-ils en conflit avec les libertés académiques? Ils sont de plus en plus utilisés pour cadrer les collaborations et sujets abordés.
- Reposted by Gaël VaroquauxInvestigations from @gaelvaroquaux.bsky.social into which procedure selects the most trustworthy predictive model to explain the effect of an intervention and support decision-making. How to select predictive models for decision-making or causal inference doi.org/10.1093/giga...
- Reposted by Gaël VaroquauxPeople think that data scientists are replacing them with AI. The scrub team is fighting back by replacing data scientists and all their practical knowledge with Python classes.
- 🚀⚡ Release: 0.5.3 Check out the release notes: skrub-data.org/stable/CHANG... Highlights below ⤵️
- Reposted by Gaël Varoquaux🧹 The Cleaner now sanitizes dataframes for the early data exploration stages
- Reposted by Gaël Varoquaux🗒️ Various improvements for the TableReport, including dark mode 🌠, Pearson's correlation for numerical values 📊, and large speed-ups for tables with many columns 🏎️
- Reposted by Gaël Varoquaux⌚ The DatetimeEncoder now adds periodic features to datetime columns, using trigonometric functions or B-splines
- Reposted by Gaël Varoquaux🚀⚡ Release: 0.5.3 Check out the release notes: skrub-data.org/stable/CHANG... Highlights below ⤵️
- Reposted by Gaël Varoquaux🗒️ Do you need to prepare a ML model, and you are working with text and strings? Skrub provides four encoders to convert strings into numerical features. 🤗 models included! What's the best? Check out our blog post to find out 👀 skrub-data.org/skrub-materi...
- Reposted by Gaël VaroquauxThe Helmholtz-ELLIS Workshop on Foundation Models in Science brought top researchers together to explore AI’s impact on e.g. materials science, genomics, and astronomy, fostering collaboration & deep discussions on key challenges like benchmarking, evaluation, and ethics. ellis.eu/news/the-hel...
- Open source is draining. One's todo-list is open to world. People seldom realize the cost of what they get for free. Unpleasant comments do happen. Cost of maintenance is not understood. opensource.com/article/17/2...
- Reposted by Gaël Varoquaux👉 Pour répondre à l'enquête : bit.ly/antony-quest... 🚆 Le thème "transports et mobilité" à été choisi par 61% des répondants jusqu'ici. 🌺 Bon week-end à tous et toutes !
- Reposted by Gaël Varoquaux
Here are six examples of how foreign visitors to the US with no criminal record are being treated: 1. Mahmoud Khalil, a green card holding student with no criminal record married to an American is abducted by ICE and is still in detention over a week later. www.nytimes.com/2025/03/19/n... ...... - 🎓Paper time!✨ #ICLR spotlight. Concluding of 5 years of research on missing values handling for prediction: Beware of diminishing returns in imputation for prediction. 1/8
