Machine Learning offers new opportunities for Marine Sciences, both for analysis large marine datasets (e.g., satellite, imaging, acoustic) and for improving marine ecosystem models. The online FORMAL mini-conference on Machine Learning for Marine Sciences (ML4MS) will showcase invited presentations on recent developments in Machine Learning for Marine Sciences and, more generally, for Environmental Sciences. It is supported at Sorbonne University by ISCD and SCAI.
Note that registration is mandatory to receive the connexion information. Please register HERE.
NEW: Link to the presentations: here.
Practical information
- Mandatory registration HERE (to receive the connection link!)
- The presentations will be given in English.
- For technical reasons, the number of participants may be limited.
Program
Mon 28 March 2022, 14h-16h CET
- Ketil Malde, Ass. Prof. at the University of Bergen (Bergen, Norway), invited Prof. at SU by SCAI, ISCD and iOcean.
Machine learning and the data-driven future of marine sciences
- Ronan Fablet, Prof. IMT Atlantique, Lab-STICC, AI Chair OceaniX (Brest, France)
End-to-end and physics-informed learning for ocean dynamics
- Benjamin Deneu, PhD candidate at INRIA (Montpellier, France)
Convolutional neural networks for species distribution modelling
Fri 1st April 2022, 14h-16h CET
- Frédéric Maps, Ass. Prof. University Laval (Québec, Canada), and Piotr Pasza Storożenko research engineer at Appsilon (Poland).
ML for plankton images: from classification to the estimation of functional traits
- Tanya Strydom, PhD candidate at Université de Montréal (Montreal, Canada)
Making something out of nothing at all: transfer learning for network prediction
- Alba Ordoñez, Senior Research Scientist at the Norwegian Computing Centre (Oslo, Norway)
Improving marine acoustic target classification with spatial information
- Maike Sonnewald, Ass. Res. Scholar at Princeton University and NOAA (Seattle, USA), and Redouane Lguensat, IRD Research Engineer at the IPSL climate modeling center in Paris, France (Recorded presentation from Maike).
Using deep learning for prediction of physical regimes and biological provinces
Organising committee
- Sakina-Dorothée Ayata (Sorbonne University, France)
- Frédéric Maps (Université Laval, Canada)
- Oliver Andrews (University of Bristol, United-Kingdom)
- Sünnje Basedow (University of Tromsø, Norway)
- Jean-Olivier Irisson (Sorbonne University, France)
About FORMAL
FORMAL aims to study the dynamics of life in the ocean at various spatial and temporal scales. It is a junior projet-team sponsored by the Institut des Sciences du Calcul et des Données (ISCD; http://iscd.sorbonne-universite.fr/) of Sorbonne Université (SU). FORMAL gathers members of various laboratories from SU (LOCEAN, LOV, LJLL, LBI2M, ISYEB, AD2M, LOMIC). To stay informed of our future activities, please register here.
