6 IRIT papers selected for the NeurIPS 2021 conference
NeurIPS (Neural Information Processing Systems) will take place from December 6 to 14, 2021. This conference is the most prestigious one organized around machine learning. It will be held remotely. This year, 6 papers co-authored by IRIT researchers have been selected by the reading committee of the conference.
Machine learning is a branch of artificial intelligence that provides computer systems with learning capabilities from data and through training models and algorithm design.
The 6 papers featured are on the following themes :
- Nonsmooth Implicit Differentiation for Machine-Learning and Optimization – Jérôme Bolte · Tam Le · Edouard Pauwels (équipe ADRIA) · Tony Silveti-Falls
- Numerical influence of ReLU’(0) on backpropagation – David Bertoin · Jérôme Bolte · Sébastien Gerchinovitz · Edouard Pauwels
- A novel notion of barycenter for probability distributions based on optimal weak mass transport – Elsa Cazelles (équipe SC)· Felipe Tobar · Joaquin Fontbona
- A PAC-Bayes Analysis of Adversarial Robustness – Paul Viallard · Eric Guillaume VIDOT (équipe ARGOS) · Amaury Habrard · Emilie Morvant
- Unbalanced Optimal Transport through Non-negative Penalized Linear Regression – Laetitia Chapel · Rémi Flamary · Haoran Wu · Cédric Févotte (équipe SC) · Gilles Gasso
- Semialgebraic Representation of Monotone Deep Equilibrium Models and Applications to Certification – Tong Chen · Jean Lasserre · Victor Magron · Edouard Pauwels
Founded in 1987, the conference is now an annual interdisciplinary and multidisciplinary meeting that includes invited talks, demonstrations, symposia, and oral and poster presentations of refereed papers. The conference is accompanied by a professional exhibition focused on machine learning in practice, a series of tutorials, and thematic workshops that provide a less formal setting for the exchange of ideas.