hugo Schmutz

Postdoc - Aix-Marseille Université

[hugo Schmutz @HugoSchmutz2]

@pamattei @3IAcotedazur @InriaMaasai @inria_sophia (8/8)

[hugo Schmutz @HugoSchmutz2]

Excited to present my article on safe semi-supervised learning at #ICLR2023 (Poster session 4, Tue 2, 4:30 pm)! Even though SSL received a lot of attention in the past years, most methods do not benefit from theoretical guarantees. (1/8) #SemiSupervisedLearning

[hugo Schmutz @HugoSchmutz2]

Additionally, we provide a theoretical explanation of the intuition of the popular SSL methods. Our results have important implications for the use of unlabelled data in deep learning. (7/8) #UnlabelledData

[hugo Schmutz @HugoSchmutz2]

Super happy to be listed as top reviewers at #aistats23 for the first review of my life! @aistats2023

[hugo Schmutz @HugoSchmutz2]

@felix_led I think Mattermost is all of the above ☝️

[hugo Schmutz @HugoSchmutz2]

Getting ready for my talk! (in french)

[hugo Schmutz @HugoSchmutz2]

Very proud to have presented my work highlighted at #EANM22 on the prediction of the lung cancer response to immunotherapy. We show that heterogeneous biomarkers can be combined to improve predictive performance. @officialEANM @Lacassagne_Nice @3IAcotedazur @InriaMaasai

[Fanny Orlhac @FOrlhac]

Development and validation of a #radiomics model for the diagnosis of dopaminergic denervation on [18F]FDOPA PET/CT just published in #EJNMMI
https://t.co/eJX1ywgrql

Congratulations to the Nuclear Medicine Department of the @Lacassagne_Nice, @Univ_CotedAzur !

[Pierre-Alexandre Mattei @pamattei]

Day 4 of @StatFr’s #Statlearn2022 was a nice sunny day of lectures by @tbmurphy on statistical models for graphs, from Erdös-Renyi to stochastic block models.

[Pierre-Alexandre Mattei @pamattei]

Day 3 of @StatFr’s #Statlearn2022 at @IES_Cargese was about double descent (of the test risk, as demystified by Claire Boyer) in the morning and double decent (of the hills of the calanche di Piana) in the afternoon 🏔⛰

[Federico Bergamin @fedebergamin]

Interested in out-of-distribution detection? Tomorrow we present our paper “Model-agnostic out-of-distribution detection using combined statistical tests” at #AISTATS2022.

March 30th, 12:30-14:00am CEST

📄https://t.co/xVWlQIzQua
🧑‍💻https://t.co/6LWm5IlJ7x

[Stat.ML Papers @StatMLPapers]

Don't fear the unlabelled: safe deep semi-supervised learning via simple debiaising. (arXiv:2203.07512v1 [https://t.co/zjV5HgYw5a]) https://t.co/MGedEI7NPT

[hugo Schmutz @HugoSchmutz2]

Additionally, a cool infographic about the environmental impact of numeric.
https://t.co/2NDRFcfMcM

[hugo Schmutz @HugoSchmutz2]

Want to test your environmental impact? Make the test on https://t.co/iAH4NPO1qR.
A very interesting publication about the environmental impact of individual actions.
https://t.co/kQEegkGQGU

[Académie des sciences @AcadSciences]

🔴En direct de la cérémonie de remise des #prixAcadSciences 2021
🏅Serena VILLATA (@serena_villata), lauréate du Prix @Inria – Académie des sciences jeunes chercheurs et jeunes chercheuses
ℹ️ https://t.co/4QaRdkl18x
👉https://t.co/z3FulEcosV