Matteo Zecchin

[Matteo Zecchin @MatteoZecchin]

If you are interested, check out the paper here: https://t.co/FS7CVhDBJI
and the blog post here: https://t.co/bIeFChmB4z

Special thanks to my exceptional co-authors, Sangwoo Park and Osvaldo Simeone, and to the CENTRIC Project for the support! (3/3)

[Matteo Zecchin @MatteoZecchin]

Excited to share that our paper “Adaptive Learn-then-Test: Statistically Valid and Efficient Hyperparameter Selection” has been accepted as a Spotlight at #ICML2025 (top 2.6%)! 🏆 (1/3)

[Matteo Zecchin @MatteoZecchin]

We introduce aLTT, a new method for hyperparameter selection that adaptively allocates testing resources—saving time & cost without losing statistical guarantees. From safe RL to prompt engineering, it enables data-efficient, reliable AI system configuration. (2/3)

[arXiv math.IT Information Theory @mathITbot]

Matteo Zecchin, Kai Yu, Osvaldo Simeone: In-Context Learning for MIMO Equalization Using Transformer-Based Sequence Models https://t.co/p7Ois7KpPv https://t.co/aKdhfe4iRM

[arXiv math.IT Information Theory @mathITbot]

Matteo Zecchin, Sangwoo Park, Osvaldo Simeone: Forking Uncertainties: Reliable Prediction and Model Predictive Control with Sequence Models via Conformal Risk Control https://t.co/uhF0iE7iZ1 https://t.co/wV9xHXLFvI

[Valeriy M., PhD, MBA, CQF @predict_addict]

Federated Inference with Reliable Uncertainty Quantification over Wireless Channels via Conformal Prediction

https://t.co/IPQhM6Gcub

https://t.co/lxwSiHub9P

#conformalprediction

[Stat.ML Papers @StatMLPapers]

When is Importance Weighting Correction Needed for Covariate Shift Adaptation?. (arXiv:2303.04020v1 [https://t.co/zjV5HgYw5a]) https://t.co/QP3f42VHFJ

[Accepted papers at TMLR @TmlrPub]

Communication-Efficient Distributionally Robust Decentralized Learning

Matteo Zecchin, Marios Kountouris, David Gesbert

https://t.co/LYM4TRrvSl

#collaboratively #minimax #gradient

[IEEE Information Theory Society @ieee_itsoc]

Two 2-year Post-Doc positions to work at the intersection of machine learning and information theory at King's College London with Prof. Osvaldo Simeone https://t.co/T9QolGWICx

[Deep RL @deep_rl]

Robust Bayesian Learning for Reliable Wireless AI: Framework and
Applications - Matteo Zecchin https://t.co/P8RUrdw86C

[午後のarXiv @arxivml]

"Robust Bayesian Learning for Reliable Wireless AI: Framework and Applications",
Matteo Zecchin, Sangwoo Park, Osva…
https://t.co/2M9kJHfpd2

[午後のarXiv @arxivml]

"Communication-Efficient Distributionally Robust Decentralized Learning",
Matteo Zecchin, Marios Kountouris, David …
https://t.co/qwFbULToN6

[午後のarXiv @arxivml]

"Robust PAC$^m$: Training Ensemble Models Under Model Misspecification and Outliers",
Matteo Zecchin, Sangwoo Park,…
https://t.co/d2wy366ZGH

[arXiv Daily @Arxiv_Daily]

A Novel Look at LIDAR-aided Data-driven mmWave Beam Selection
https://t.co/Ux2msyfFDm
by Matteo Zecchin et al. including @denizgunduz1
#NeuralNetwork #LossFunction

[Windmill-ITN @ItnWindmill]

Our WindMill team of Early Stage Researchers is now almost complete. Welcome to Shirin Goshtasbpour @ETH Zürich, @majadoon at CTTC, Matteo Zecchin at EURECOM, Anay Deshpande at Università degli Studi di Padova. #MSCA #ITN