Tweet ghim
Pablo Groisman 🎲
9.4K posts
Pablo Groisman 🎲
@pgroisma
Matemático Abrazar el azar https://t.co/DUbUcizCdg Te Regalo Un Teorema https://t.co/07qzMEYmfo
Tham gia Haziran 2010
948 Đang theo dõi9.5K Người theo dõi
Y pensar que mis estudiantes se quejan de que sus simulaciones tardan mucho en correr...
Tendencias en Argentina@porqueTTarg
"Dados" Porque lanzó los dados 10.000 veces durante más de 17 horas y documentó los resultados, una distribución casi perfecta.
Español
Pablo Groisman 🎲 đã retweet
Holy shit... MIT just made their 505-page Machine Learning textbook completely free, grab it before it's paid again.
It's called "Foundations of Machine Learning" and it covers everything from PAC learning to deep learning theory.
No $150 textbook. No paywalled courses. No gatekeeping.
Here's how to get it: ↓
English
Many thanks to my amazing co-authors Matthieu Jonckheere, Jordan Serres and Mariela Sued.
English
(i) Isomap and Fermat geodesic distances on manifolds, (ii) difussion distances, (iii) Wasserstein distances computed with respect to learned ground metrics, (iv) first passage percolation and (v) discrete approximations of length spaces.
English
Very happy with our new paper
k-means with learned metrics
arxiv.org/html/2603.1460…
In many learning tasks the underlying geometric structure of the data is not directly observed but must be inferred.
English
Very proud of this group. The new generation of probabilists/statisticians/machine-learningers/mathematicians/data scientists (however you want to call them) at UBA.
This is at XVII CLAPEM, the amazing event organized by our Uruguayans brothers. Thanks to them for that.
English
Comparto info sobre escuela destinada a estudiantes avanzados de carreras de grado de matemática, ciencia de datos y afines en @FIQUNL
deptomatefiq.github.io/madatop/
Español
Very happy to join this adventure. Thanks @ch4rleston !
Carlos Sarraute@ch4rleston
🌟 It’s an honor to introduce Pablo Groisman, part of our Scientific & Industry Advisors team at @DisruptiveRI. @PGroisma is a Principal Researcher at CONICET at the Instituto de Investigaciones Matemáticas Luis A. Santaló (IMAS), and a Full Professor in @DM_ExactasUBA. He holds a PhD in Mathematics from the University of Buenos Aires. His research spans probability theory and its applications, including interacting particle systems, branching and random growth processes, stochastic models for synchronization, quasi-stationary distributions, stochastic differential equations, geometric and topological statistics, and reconstruction in dynamical systems. Pablo is deeply engaged in the scientific community. He serves as an Associate Editor of "Chaos, Solitons & Fractals", has held key academic leadership roles at UBA, and is the author of more than 50 research articles as well as several popular science books in Spanish. 🔬 Together with @AndresTobelem, @Viio_J, @MOskarsdottir and @D_E_Fernandez, Pablo brings strong theoretical depth and scientific leadership to our mission of bridging rigorous research with real-world impact. #DisruptiveResearch #AppliedMathematics #DataScience
English
This is great news! Congrats @ch4rleston !
Carlos Sarraute@ch4rleston
🚀 I’m happy to present the Disruptive Research Institute! @DisruptiveRI A new research initiative providing end-to-end solutions in Network and Data Science for private companies and public organisations — bridging academic research and real-world impact. We are supported by an outstanding team of Scientific and Industry Advisors, bringing deep expertise from academia and industry. I will be presenting them over the coming days. 👉 Follow @DisruptiveRI to stay up to date with our latest work and projects. #NetworkScience #DataScience #AppliedResearch #DisruptiveResearch
English
@wisdom_in_op @arxiv @ianbounos37 Gracias! Todavía estamos en etapa exploratoria con las aplicaciones concretas. Si tenés algo en mente, nos encataría escucharte.
Español
@pgroisma @arxiv @ianbounos37 Enhorabuena por el paper. La solución de forma cerrada para extracción lineal es elegante. ¿Habéis considerado aplicaciones en ingeniería neural? #neurociencia
Español
Very happy that we posted to @arxiv our "Christmas paper"
Invariant Feature Extraction Through Conditional Independence and the Optimal Transport Barycenter Problem: the Gaussian case
with @ianbounos37, Mariela Sued and Esteban Tabak
arxiv.org/abs/2512.20914
English
I am very proud and grateful to be part of this wonderful team. I hope you enjoy the article. We welcome your comments.
Happy New Year!
English
We tackle this problem with the tools provided by the optimal transport problem and the Wasserstein barycenter.
English