NuTS first workshop

Welcome to the Réseau Numérique en Terre Solide (NuTS), a CNRS thematic network (RT) that aims to foster the community of developers and users of numerical tools for solid earth science. Our goal is to make this community more visible, efficient, and informed through a series of conferences and training courses on digital science, from data manipulation to intensive scientific computing.

We are excited to announce the first workshop of the RT NuTS, which will take place from May 30th to June 2nd in Lyon. This workshop is designed to introduce computing resources and machine learning techniques to our community, and it is divided into two parts: the first two days will feature scientific talks, and the last two days will offer hands-on practicals on machine learning.

During the practicals, attendees will have the opportunity to work with Jupyter Notebooks and Python, using the cutting-edge computing resources provided by the GLiCID mesocentre, including a series of A100 GPU. The practicals will cover a range of topics, including data preparation, dimensionality reduction, classification and regression with Scikit-Learn, and deep learning with Pytorch. By the end of the workshop, attendees will have gained practical experience with these tools and techniques, and will be better equipped to apply machine learning to their own research in solid earth science.

We are planning a poster session dedicated to AI-related works, particularly those that are in progress and require further discussion. We welcome contributions from students and researchers who are interested in sharing their plans and unfinished work with the community.

We invite you to join us for this exciting workshop, where you will learn from experts in the field, gain hands-on experience with machine learning tools, and network with other researchers in solid earth science. Don't miss this opportunity to advance your skills and knowledge in digital science and machine learning.

To register for the workshop, please click the "Register Now" button. If you have any questions or need assistance with registration, please don't hesitate to contact us. We look forward to seeing you in Lyon!

The registrations are now closed

Event location:

Salles Fontanes, Batiment Darwin
3-9 Rue Raphaël Dubois, 69100 Villeurbanne

More details on how to get there can be found here.

Program

Mardi 30 Mai

13h30 Accueil - Yann Capdeville (LPG, CNRS, NuTS)

13h45 Yann Capdeville (LPG, CNRS, NuTS) - Hardware and Ressources for scientific HPC

14h45 Poster pitches

Gautier Laurent (ISTO)- Geocognitive knowledge-based geological modelling : proof of concept

Céline Hourcade (LPG) - Detection and discrimination of low magnitude seismic events: Application to the Armoricain Massif

Buchanan Kerswell (Géosciences Montpellier) - Computing Rates and Distributions of Rock Recovery in Subduction Zones

Théo Santos (LGLTPE/CRAL) - Generative neural networks for downscaling the upper mantle

Amandine Fratani (RING, GeoRessources) - Can neural networks learn how to associate fault data? First results

Mathieu Nougaret (IPGP) - The use of machine learning algorithms to help in the prediction of eruption at the Piton de la Fournaise volcano

Said Sadeg (GeoRessources) - Analyse des données de la μ-XRF

Nicolas Gilardi (BRGM) - Geoscientific benchmarks for AI methods

Yuqing Xie (Géoazur) - Deep Learning-based array processing with RNN and MLP

15h30 Pause

16h00 - 18h00

Hatim Bourfoune (IDRIS) - Jean Zay

Aurélien Garivier (ENS Lyon) - Understanding the Efficiency of Machine Learning: Progress and Challenges

Mercredi 31 Mai

9h00-10h30

Maëlis Arnould (LGLTPE) - Diamond open access journals

Marielle Malfante (CEA) - AI for monitoring: focus on reliability and anomaly detection

David Michéa (EOST)-  Automatic machine learning classification of ground displacement data cubes: application to large InSAR and optical datasets

10h30 Pause, Group photo

11h00-12h30

Sophie Giffard (ISTerre) - Examples and best practices of using machine learning in geosciences, especially neural convolution networks

Damia Benet (Earth Observatory of Singapore)- Machine learning for volcanic ash studies

Thomas Bodin (LGLTPE) - Using Generative Networks for Inverse problems

12h30-14h00 Repas et Posters

14h00-15h30

Nestor Cerpa (Géosciences Montpellier) - Using deep learning to predict the evolution of mechanical anisotropy in the Earth's mantle due to texture development

Nathanael Schaeffer (ISTerre) - Neural networks and inverse problem: application to Earth's core flow estimation.

Léonard Seydoux (IPGP) - AI-based seismic and geodetic data fusion for understanding geophysical phenomena.

15h30 Pause

16h00-17h30

Gautier Laurent (ISTO) - A Geocognitive Approach to Machine Interpretation of Geological Structures

Clément Hibert (ITES) - Contribution of machine learning to environmental seismology

Quentin Blétery (Géoazur)- Can AI anticipate earthquakes?

17h30-18h00 Table ronde

Evening: conference dinner at 19h30. Brasserie des Brotteaux, 1, place Jules Ferry, Lyon.

Jeudi 1er Juin


Matin (9h-12h00): Atelier 1 - First steps with data preparation

Here we will investigate the basic statistical properties and some feeling of the data. Do we have sufficient training points? Is the data stationary? What features will be relevant to solve the task at hand? Can we avoid overfitting? 
 
Coordinateur: Léonard Seydoux (IPGP)

12h00-14h00 Repas et Posters

Après midi (14h00-18h00): Atelier 2. Machine learning approches

If deep learning is powerful to solve a vast majority of problems, machine-learning approches, if successful, provide more insight about the physics at play for solving problems. We will investigate and criticize several supervised and unsupervised approaches to solve the task of classifying clouds of Lidar points (scenes) to infer the type of structure visible therein.

Coordinateur: Léonard Seydoux (IPGP)

Vendredi 2 Juin

Matin: (9h00-12h00) Atelier 3. Deep Learning with Pytorch

When the attempts to solve the problem are unsuccessful or if defining features is a challenge because of the input data's dimension, we can also learn the features that best solve the task. Here we will revisit the problem of Lidar with deep-learning approaches and see how we can efficiently design a neural network to solve the classification task. 

Coordinateur :  Léonard Seydoux (IPGP)

12h00-14h00 Repas et Posters

Après midi (14h00-15h30): Atelier 4. Deeper dive in artificial intelligence

This last class will be a free hands class on various problems. We invite people to bring datasets and tasks of interest so we can sit and try machine-learning solutions for solving them. We will also dive deeper into algorithms, try to improve performances and discuss some good practices in machine and deep learning. 
 
Coordinateur : Léonard Seydoux (IPGP)

 

 

 

 

 

Online user: 2 Privacy
Loading...