Talks and presentations

Most slides are made with Powerpoint, but the given links usually give access to pdf versions which are lighter but may contain small glitches or low-quality images. Contact me if you want the .pptx version. You are more than welcome to re-use some illustrations from my slides.

Thesis defense: Sketching for Large-Scale Learning of Mixture Models (Oct. 2017) [Slides]

Seeking universal approximation for continuous limits of GNNs on large random graphs . Oct. 2024 Slides
- Asilomar 2024: Understanding Graph Neural Networks

Graph Neural Networks on Large Random Graphs: Convergence, Stability, Universality . 2021-2024 Slides
- ICIAM 2023: Mini-symposium on Geometric methods in machine learning and data analysis
- Journées Géosto MIA 2022
- Séminaires MLSP, ENS Lyon
- SIAM Imaging Sciences 2022: Early Career Prize Lecture
- Journée MAS 2020
- GdR MIA Day: High-Dimensional Data Analysis
- Séminaires LJAD (Nice)
- SIAM/IS Early Career Prize
- ML-MTP : Machine Learning in Montpellier, Theory and Practice
- Séminaire ENS Lyon, DANTE

Entropic Optimal Transport and Wasserstein Barycenters in Random Graphs . Nov. 2021 - Mar. 2023 Slides
- GdR ISIS - MIA day: Optimal Transport and Statistical Learning
- One World MINDS Seminar

Not too little, not too much: a theoretical analysis of graph (over)smoothing . Dec. 2022 Slides
- NeurIPS22 (Oral)
- Learning on Graphs Conference 22 (Spotlight)

Graph Neural Networks: Introduction, some theoretical properties . Jan. 2021 Slides
- Data Science Master Seminar, Lille
- GAIA team seminar

Convergence and Stability of GCNs on Large Random Graphs . Dec. 2020 Slides
NeurIPS (Spotlight)

Sparse and Smooth: Spectral Clustering in the dynamic SBM . Dec. 2020 Slides
iTWIST 2020

A short introduction to graphon . Mar. 2020 Slides
GAIA team seminar

Universal Invariant and Equivariant Graph Neural Networks . May. 2019 - Mar. 2020 Slides
- LMS-Bath Symposium: Mathematics of Machine Learning, poster
- ELLIS Workshop on Geometric and Relational Deep Learning
- Journée du GdR Isis : Graph Signal Processing
- Journée du GdR Isis : Theory of Deep Learning
- NeurIPS 2019, Vancouver

Exponentially smoothed spectral clustering and dynamic stochastic block model . May. 2019 Slides
Graph signals : learning and optimization perspectives, Université de Montpellier, France.

Scalable, online change point detection with NEWMA . Nov. 2018 - Jan. 2019 Slides
- CICS team seminar, GIPSA-lab, Grenoble
- SELECT team seminar, INRIA Saclay, Orsay
- Séminaire Probabilités et Statistiques, Laboratoire Paul Painlevé, Lille
- Séminaire Image Optimisation et Probabilités, Institut de Mathématiques de Bordeaux
- CERMICS, ENPC, Marne-la-vallée
- Equipe DANTE, ENS Lyon

Fisher metric, support stability and optimal number of measurements in compressive off the grid recovery . Oct. 2018 - Nov. 2018 Slides
- Conférence OMS, Institut de Mathématiques de Toulouse, Toulouse
- Séminaire S3, CentraleSupélec, L2S, Paris
- Séminaire de l'équipe Sigma, CRIStAL, Lille
- Ecole Normale Supérieure de Lyon, Laboratoire de Physique, Lyon - THOTH team seminar, INRIA Rhône-Alpes, Grenoble

Sketched Learning from Random Feature Moments . International Symposium on Math. Programming (ISMP), July 2018, Bordeaux, France. Slides

NEWMA: a new method for scalable model-free online change-point detection . Séminaire SPOC, June 2018, Dijon, France. Slides

A Dual Certificate Analysis of Compressive Off-the-grid Recovery . Journee thématique du GDR MIA: Parcimonie et applications, IMB, May 2018, Talence, France. Slides

Sketched Learning from Random Features Moments . Telecom-Paristech, Apr. 2018, Paris. Slides

Online sketches with random features . Paris Science and Data, IPGG, Apr. 2018, Paris. Slides

Sketched Learning from Random Features Moments . Imaging in Paris, IHP, Apr. 2018, Paris. Slides

Sketching for Large-Scale Learning of Mixture model . Nov. 2017 - Dec 2017. Slides
- Gatsby Computational Neuroscience UCL, London
- CFM-ENS Data Science chair Annual Conference, ENS, Paris
- The Future of Random Projections : a mini-workshop, IPGG, Paris. Video

Spike super-resolution with random Fourier sampling . SPARS, Jun. 2017, Lisbon. Slides, Abstract

Random Moments for Sketched Mixture Learning . SPARS, Jun. 2017, Lisbon. Slides, Abstract

Best Student Paper Award, SPARS 2017.

Sketching for Large-Scale Learning of Mixture model . Jan. 2017 - Mar. 2017 Slides
- Thesis pre-defense, IRISA, Rennes
- Séminaires de Statistiques de l’IRMAR, IRMAR, Rennes
- IdCOM Seminars, University of Edinburgh, Edinburgh
- Journée Classification MimeTIC, IRISA, Rennes.

Sketching for Large-Scale Learning of Mixture model . Ecole d'été Peyresq, Peyresq, Jul. 2016. Slides

Sketching for Large-Scale Learning of Mixture model . Journée du GdR Isis: Algorithmes gloutons pour l'optimisation sous contrainte de parcimonie, Paris, Jun. 2016. Slides

Sketching for Large-Scale Learning of Mixture model . ICASSP, Shanghaï, Mar. 2016.

Sketching for Large-Scale Learning of Mixture model . International Matheon Conference on Compressed Sensing and its Applications, Dec. 2015, Berlin.

Compressive Gaussian Mixture Estimation by Orthogonal Matching Pursuit with Replacement . SPARS, Jul. 2015, Cambridge.

Nicolas Keriven

Talks and presentations