Machine Learning for Radio Weak Lensing

Machine Learning for Radio Weak Lensing

Position: PhD
Deadline:  22/04/2021
Contact: Samuel Farrens, Valeria Pettorino

Details about this position are provided on ADUM. [PDF]

This project proposed in the context of the Data Intensive Artificial Intelligence (DIAI) call, published by the data intelligence institute of Paris (diiP).

PhD scholarships on Data Intensive Artificial Intelligence (DIAI)

Simulation-based cosmological inference of large galaxy surveys

Simulation-based cosmological inference of large galaxy surveys

Position: PhD
Deadline:  15/04/2021
Contact: Martin Kilbinger, Valeria Pettorino, François Lanusse

Details about this position are provided in the following pdf file.

Please apply here on adum.fr, the central site to apply for this PhD.

This project proposed in the context of UDOPIA, the Université Paris-Saclay Doctoral Programme in Artificial Intelligence

Dark Energy tomography with the Euclid survey

PhD topic on Dark Energy tomography with the Euclid satellite

Position: PhD
Deadline:  15/04/2021
Contact: Valeria Pettorino

Details about this position are provided in the following PDF.

Interested candidates should send a CV and exam record to Valeria Pettorino, and arrange for 2 reference letters to be sent separately. The application e-mail should be preferably in English. Knowledge of cosmology, general relativity or previous use of CAMB/CLASS codes are an advantage. 
Projects are not funded yet and decision on fundings is expected for Spring 2021. In the meantime, shortlisted candidates will be asked to register to ADUM https://www.adum.fr/. 

“3x2pt'” analysis: Cross-correlations of cosmological probes, and application to state-of-the-art weak-lensing and galaxy clustering surveys.

“3x2pt'” analysis: Cross-correlations of cosmological probes, and application to state-of-the-art weak-lensing and galaxy clustering surveys.

Position: PhD
Deadline:  15/04/2021
Contact: Martin Kilbinger, Valeria Pettorino

Details about this position are provided in the following PDF.

Deep Learning for space-variant deconvolution in galaxy surveys

 

Authors: Florent Sureau, Alexis Lechat, J-L. Starck
Journal: Astronomy and Astrophysics
Year: 2020
DOI: 10.1051/0004-6361/201937039
Download: ADS | arXiv


Abstract

The deconvolution of large survey images with millions of galaxies requires developing a new generation of methods that can take a space-variant point spread function into account. These methods have also to be accurate and fast. We investigate how deep learning might be used to perform this task. We employed a U-net deep neural network architecture to learn parameters that were adapted for galaxy image processing in a supervised setting and studied two deconvolution strategies. The first approach is a post-processing of a mere Tikhonov deconvolution with closed-form solution, and the second approach is an iterative deconvolution framework based on the alternating direction method of multipliers (ADMM). Our numerical results based on GREAT3 simulations with realistic galaxy images and point spread functions show that our two approaches outperform standard techniques that are based on convex optimization, whether assessed in galaxy image reconstruction or shape recovery. The approach based on a Tikhonov deconvolution leads to the most accurate results, except for ellipticity errors at high signal-to-noise ratio. The ADMM approach performs slightly better in this case. Considering that the Tikhonov approach is also more computation-time efficient in processing a large number of galaxies, we recommend this approach in this scenario.

In the spirit of reproducible research, the codes will be made freely available on the CosmoStat website (http://www.cosmostat.org). The testing datasets will also be provided to repeat the experiments performed in this paper.