Associate Prof at University of Novi Sad & researcher at Istituto Italiano di Tecnologia. Sci adventures include NumLinAlg, Optim, ML, OT,…
Finale Ligure, Italy
Joined April 2010
- Tweets762
- Following357
- Followers113
- Likes5.8K
Vladimir Kostic retweeted
Dr MMD is at #ICML2026!
(De)-regularized Maximum Mean Discrepancy Gradient Flow
HALL A #4302
Wed, Jul 8, 2026 • 2:30 PM – 4:15 PM
@Hudson19990518 Mustafi @PierreGlaser @Korba_Anna Sriperumbudur
Vladimir Kostic retweeted
Outcome Aware Spectral Feature Learning for Instrumental Variable Regression, at #ICML2026
HALL A #4413
Tue, Jul 7, 2026 • 10:30 AM – 12:15 PM
@DimitriMeunier1 Wornbard @vkostic30 @antoine_mln Fröhlich Lounici @MPontil
Vladimir Kostic retweeted
Can contrastive learning produce equivariant representations with statistical guarantees for downstream regression, conditional prob estimation, and uncertainty quantification?
Come discuss eNCP at #ICML2026
Poster: Wed Jul 8, 5:00–6:45 PM
Paper/Code: bit.ly/4gj9yUE
Vladimir Kostic retweeted
We will free women n children in Afg - Kills 3.2 million
We will free women n children in Iraq - Kills 1.9 million
We will free women n children in Libya - Kills 2.6 million
We will free women n children in Iran - Kills 100 girls in 30 minutes
Saviours of women
Vladimir Kostic retweeted
At #NeurIPS ? Visit our posters! 🧵
Demystifying Spectral Feature Learning for Instrumental Variable Regression: #2600, Wed 11am
Regularized least squares learning with heavy-tailed noise is minimax optimal: #3012, Wed 4:30pm ✨spotlight✨
1/2
Vladimir Kostic retweeted
Online talk on Wednesday - looking forward to it!
How can we tell whether an action really causes a result and not just coincides with it?
Don't miss out on this #AIforGood Discovery session with @ArthurGretton (@ucl) and moderator @Conrad__Philipp (@UniJena) to explore how #AI and causal modeling can help reveal the true impact of interventions – from social programs and medical treatments to online policies – even when experiments aren’t possible.
Register now:
📅 29 October 2025
🕔 17:00–18:00 CEST
🌐 loom.ly/fqTcQMk
Vladimir Kostic retweeted
Distributional diffusion models with scoring rules at #icml25
Fewer, larger denoising steps using distributional losses!
Wednesday 11am poster E-1910
arxiv.org/pdf/2502.02483
@agalashov
@ValentinDeBort1
Guntupalli
@zhouguangyao
@sirbayes
@ArnaudDoucet1
Vladimir Kostic retweeted
🚨 Poster at #AISTATS2025 tomorrow!
📍Poster Session 1 #125
We present a new empirical Bernstein inequality for Hilbert space-valued random processes—relevant for dependent, even non-stationary data.
w/ Andreas Maurer, @vkostic30 & @MPontil
📄 Paper: openreview.net/forum?id=a0EX…
Vladimir Kostic retweeted
For the past four years, I’ve been working on a topic that’s both fascinating and challenging to explain. In this post, I’ve tried to present The Operator Way — a paradigm for understanding dynamical processes — in plain, approachable terms.
pietronvll.github.io/the-ope…
Vladimir Kostic retweeted
*PLS SHARE*
Open position for an *Associate Professor* in Machine Learning at our department (@enginyeria_upf / @UPFBarcelona), via the Serra Hunter programme.
DEADLINE: January 13th 2025
upf.edu/web/personal/ofertes…
Short anouncment: for all those into #numerical linear algebra and/or stats and/or #machinelearning and #AI in numerical #computations, checkout the conference I am co-organizing in #NoviSad in #Serbia.
ala2025.pmf.uns.ac.rs
Happy and grateful be part of this great bunch of ppl! Thank you all for shared joy and passion across different projects 😃😃😃
At #NeurIPS2024 🇨🇦 our group will present 7 contributions! These span a diverse array of topics: from theoretical advances in stochastic processes and reinforcement learning to applications in molecular dynamics and uncertainty quantification.
Vladimir Kostic retweeted
At #NeurIPS2024 🇨🇦 our group will present 7 contributions! These span a diverse array of topics: from theoretical advances in stochastic processes and reinforcement learning to applications in molecular dynamics and uncertainty quantification.
Vladimir Kostic retweeted
I Will be presenting our work with @vkostic30 @GroupParrinello and @MPontil using the infinitesimal generator(IG) to recover dynamical properties of a molecular system from biased simulations
⚙️ From Biased to Unbiased Dynamics: An Infinitesimal Generator Approach
✍️@TDevergne, @vkostic30, @GroupParrinello, @MPontil
📃arxiv.org/abs/2406.09028 | 📅 Poster #3806 Dec 12, 16:30 - 19:30.
Vladimir Kostic retweeted
We often think of an "equilibrium" as something standing still, like a scale in perfect balance.
But many equilibria are dynamic, like a flowing river which is never changing—yet never standing still.
These dynamic equilibria are nicely described by so-called "detailed balance"
