Computational Statistics and Machine Learning Lab @IITtalk | PI: @MPontil | Statistical learning theory, ML for dynamical systems, ML for science, optimization.
1/ ๐ Over the past two years, our team, CSML, at IIT, has made significant strides in the data-driven modeling of dynamical systems. Curious about how we use advanced operator-based techniques to tackle real-world challenges? Letโs dive in! ๐งต๐
๐ข Great News: Congratulations to our PhD student, @AlekFrohlich, on being selected as a 2026 Google PhD Fellow in ML & Foundations! ๐
This fellowship recognizes Alek's research on representation learning and causal inference.
We're very proud of you, Alek. Well deserved!
Check out some of our recent papers on this topic and the nice blog posts by Vladimir Kostic and Pietro Novelli to get a flavor of what we'll do: https://t.co/yAn6wxol4o
2/2
๐ข The call is extended
We would be particularly keen to hear from researchers in the following areas:
1) data-driven dynamical systems, SDEs and PDEs;
2) mathematical statistics and machine learning theory; or
3) numerical and computational linear algebra.
1/2
๐น Job alert: Postdoctoral researcher in Machine Learning and Dynamical Systems at Italian Institute of Technology (IIT)
๐ Genoa ๐ฎ๐น
โฐ 31 October
๐ https://t.co/6MojuO6ZL0
Looking for your next step in machine learning research?
Join our team! We have an open #postdoc position in #MachineLearning & dynamical systems. Set in a vibrant research environment (with great colleagues, food, and sea views included).
Details and application link below
๐น Job alert: Postdoctoral researcher in Machine Learning and Dynamical Systems at Italian Institute of Technology (IIT)
๐ Genoa ๐ฎ๐น
โฐ 31 October
๐ https://t.co/6MojuO6ZL0
๐ข Weโre hiring!
๐ง Excited about ML foundations & dynamical systems?
We have an open Postdoc position in Massimiliano Pontil's CSML unit at @IITalk (Genoa, Italy ๐ฎ๐น) on the ERC "LEO" project!
โณ Up to 3 years
๐ Deadline: October 31, 2026
Details: ๐
https://t.co/WcGs3ScdT7
We invite 4-page, non-archival short papers, with accepted work presented as posters and selected contributions invited for talks.
๐ ย Submission deadline: August 29 (AoE)
๐ย Paris โ NeurIPS 2026, December 12 or 13
๐ย Website: https://t.co/HSBiKdQXCK
Representations for the Physical Sciences Workshop @ NeurIPS 2026๐ซ๐ท
Weโre bringing together researchers working on AI for science, representation learning, and physical systems, with a particular focus on self-supervision, transfer learning, adaptive sampling, and tokenization.
๐ Thursday, July 9 ยท 5:00โ6:45 PM KST ยท Hall A ยท Poster #4517 Generalization of Gibbs and Langevin Monte Carlo Algorithms in the Interpolation Regime
https://t.co/D5EKdXUjlp
If you're attending ICML, we'd be delighted to see you at the posters and discuss our work.
ICML 2026 is underway in Seoul, and I'm delighted that our group has four papers on the program this week:
๐ Tuesday, July 7 ยท Hall A ยท Poster #4413 Outcome-Aware Spectral Feature Learning for Instrumental Variable Regression
https://t.co/4LVRpKVPNI
๐ Wednesday, July 8 ยท 5:00โ6:45 PM KST ยท Hall A ยท Poster #4401 Representation Learning for Equivariant Inference with Guarantees
https://t.co/2XOlWr2867
Almost 5 years in the making... "Hyperparameter Optimization in Machine Learning" is finally out! ๐๐
From basic tuning to Bayesian optimization and meta-learning, we show how these techniques tie together in one self-contained guide.
Check it out:
๐ https://t.co/nkukNnkw2j
๐จ OpenReview might have leaked names, but it won't leak the best hyperparameters, unfortunately! ๐
Tired of the drama? Solve your HPO problems before the ICML deadline with this new book by our own Luca Franceschi & Massimiliano Pontil (& colleagues).
https://t.co/nkukNnkw2j
He will also present an entropy-respecting forwardโbackward learning scheme that mitigates the inherent ill-posedness of stochastic learning problems.
Join us for what promises to be a very insightful session!
๐ข Upcoming Talk at Our Lab
Weโre excited to host Arthur Bizzi from EPFL for a research talk next week!
Title: Towards Neural Kolmogorov Equations: Parallelizable SDE Learning with Neural PDEs
๐ Date: November 19
โฐ Time: 16:00 CET
๐ Galileo Sala, CHT @IITalk
In this talk, Arthur Bizzi will introduce Neural Kolmogorov Equations, a deterministic and parallelizable framework for learning continuous-time stochastic processes using Forward and Backward Kolmogorov Equations.
๐ข Now itโs out on arXiv! If youโre interested in theoretical guarantees for learning with dependent Hilbert-space data, check out our full paper:
๐ An Empirical Bernstein Inequality for Dependent Data in Hilbert Spaces and Applications
๐ https://t.co/h1Io7BykqX
Check out our new work on learning ambidexterous bi-manual (and multi-arm) manipulation via morphological symmetry exploitation with Equivariant NNs.
Kudos to @softraeh for leading this work.