École Euclid 2023 Rodolphe Clédassou

Date: August 15 - August 30, 2023

Venue: Ronce-les-Bains, France

Website: https://ecole-euclid.cnrs.fr/2023-accueil/

``Weak gravitational lensing''  (Le lentillage gravitationnel), cycle 2, Martin Kilbinger.

Find here links to the lecture notes and TD exercises.

  • Lecture notes (pdf).
  • TD: CFIS/UNIONS weak-lensing data analysis; cluster lensing.
    Please download:

``Scientific Software Development for Euclid'' , cycle 2, Samuel Farrens.

École Euclid de cosmologie 2021

Date: August 16 - August 27, 2021

Venue: Anglet, France

Website: https://ecole-euclid.cnrs.fr/2021-accueil/

Lecture ``Weak gravitational lensing''  (Le lentillage gravitationnel), cycle 2, Martin Kilbinger.

Find here links to the lecture notes, TD exercises, "tables rondes" topics, and other information.

  • Resources.
    • A great and detailed introduction to (weak) gravitational lensing are the 2005 Saas Fee lecture notes by Peter Schneider. Download Part I (Introduction to lensing) and Part III (Weak lensing) from my homepage.
    • Check out Sarah Bridle's video lectures on WL from 2014.
  • TD cycle 2, Data analysis. TBD
  • Lecture notes.

Cosmostat Day on Machine Learning in Astrophysics

Date: March the 5th, 2021

Organizer:  Joana Frontera-Pons  <joana.frontera-pons@cea.fr>

Venue: Remote conference. Zoom link:https://esade.zoom.us/j/88535176160?pwd=RzU1cHA5Z0xrWXkyN0x1a2tJSHZ1Zz09

On March the 5th, 2021, we organize the 6th day on machine learning in astrophysics at DAp, CEA Saclay. 


All talks are taking place remotely

13:30 - 13:40h. Welcome message                                                   
13:40 - 14:20h. Data-driven detection of multi-messenger transientsIftach Sadeh (Deutsches Elektronen-Synchrotron)
14:20 - 15:00h. Deep Learning in Radio AstronomyVesna Lukic (Vrije Universiteit Brussel)   
15:00 - 15:40h. Machine Learning for Galaxy Image Reconstruction with Problem Specific Loss - Fadi Nammour (CosmoStat - CEA Saclay)   

15:40 - 16:00h. Coffee break with virtual croissants

16:00 - 16:40h. Anomaly detection with generative methodsColoma Ballester (Universitat Pompeu Fabra)
16:40 - 17:20h. Deep learning for environmental sciencesJan Dirk Wegner (ETH Zurich)
17:20 - 18:00h. Graph Neural NetworksFernando Gama ( University of California, Berkeley)

18:00 - 18:05h. End of the day

Data-driven detection of multi-messenger transients

Iftach Sadeh (Deutsches Elektronen-Synchrotron)

The primary challenge in the study of explosive astrophysical transients is their detection and characterisation using multiple messengers. For this purpose, we have developed a new data-driven discovery framework, based on deep learning. We demonstrate its use for searches involving neutrinos, optical supernovae, and gamma rays. We show that we can match or substantially improve upon the performance of state-of-the-art techniques, while significantly minimising the dependence on modelling and on instrument characterisation. Particularly, our approach is intended for near- and real-time analyses, which are essential for effective follow-up of detections. Our algorithm is designed to combine a range of instruments and types of input data, representing different messengers, physical regimes, and temporal scales. The methodology is optimised for agnostic searches of unexpected phenomena, and has the potential to substantially enhance their discovery prospects.

Deep Learning in Radio Astronomy

Vesna Lukic (Vrije Universiteit Brussel)

Machine learning techniques have proven to be increasingly useful in astronomical applications over the last few years, for example in image classification and time series analysis. A topic of current interest is the classification of radio galaxy morphologies, as it gives us insight into the nature of the Active Galactic Nuclei and structure formation. Future surveys such as the Square Kilometre Array (SKA), will detect many million sources and will require the use of automated techniques. Convolutional neural networks are a machine learning technique that have been very successful in image classification, due to their ability to capture high-dimensional features in the data. We show the performance of simple convolutional network architectures in classifying radio sources from the Radio Galaxy Zoo. The use of pooling in such networks results in information losses which adversely affect the classification performance, however Capsule networks preserve this information with the use of dynamic routing. We explore a couple of convolutional neural network architectures against variations of Capsule network setups and evaluate their performance in replicating the classifications of radio galaxies detected by the Low Frequency Array (LOFAR). Finally, we also show how it is possible to use convolutional neural networks to find sources in radio surveys.

Machine Learning for Galaxy Image Reconstruction with Problem Specific Loss

Fadi Nammour (CosmoStat - CEA Saclay)

Telescope images are corrupted with blur and noise. Generally, blur is represented by a convolution with a Point Spread Function and noise is modelled as Additive Gaussian Noise. Restoring galaxy images from the observations is an inverse problem that is ill-posed and specifically ill-conditioned. The majority of the standard reconstruction methods minimise the Mean Square Error to reconstruct images, without any guarantee that the shape objects contained in the data (e.g. galaxies) is preserved. Here we introduce a shape constraint, exhibit its properties and show how it preserves galaxy shapes when combined to Machine Learning reconstruction algorithms.

Anomaly detection with generative methods

Coloma Ballester (Universitat Pompeu Fabra)

Anomaly detection is frequently approached as out-of-distribution or outlier detection. In this talk, a method for out-of-distribution will be discussed. It leverages the learning of the probability distribution of normal data through generative adversarial networks while simultaneously keeping track of the states of the learning to finally estimate an efficient anomaly detector.

Deep learning for environmental sciences

 Jan Dirk Wegner (ETH Zurich) 

A multitude of different sensors is capturing massive amounts of geo-coded data with different spatial resolution, temporal frequency, viewpoint, and quality every day. Modelling functional relationships for applications is often hard and loses predictive power due to the high variance in sensor modality. Data-driven approaches, especially modern deep learning, come to the rescue and learn expressive models directly from (labeled) input data. In this talk, I will present deep learning methods to analyze geospatial data at large scale for two specific applications in the environmental sciences: biodiversity estimation and global vegetation height mapping.

Graph Neural Networks

Fernando Gama ( University of California, Berkeley)

Graphs are generic models of signal structure that can help to learn in several practical problems. To learn from graph data, we need scalable architectures that can be trained on moderate dataset sizes and that can be implemented distributedly. In this talk, I will draw from graph signal processing to define graph convolutions, and use them to introduce graph neural networks (GNNs). I will prove that GNNs are permutation equivariant and stable to perturbations of the graph, properties that explain their scalability and transferability. I will also use these results to explain the advantages of GNNs over linear graph filters. I will then discuss the problem of learning decentralized controllers, and how GNNs naturally leverage the partial information structure inherent to distributed systems. Using flocking as an illustrative example, I will show that GNNs, not only successfully learn distributed actions that coordinate the team but also transfer and scale to larger teams.

 Previous Cosmostat Days on Machine Learning in Astrophysics :

CosmosClub: Ariel Sánchez (09/07/20)

CosmosClub Ariel Sánchez

Date: July 9th 2020, 10.00 a.m.

Speaker: Ariel Sánchez (MPE Garching/ Max-Planck-Institut für extraterrestrische Physik )

Title:   Let us bury the prehistoric h: arguments against using h1Mpc units in observational cosmology

Room: Zoom Meeting (connection details will be updated soon)


It is common to express cosmological measurements in units of h^-1 Mpc. Here, we review some of the complications that originate from this practice. A crucial problem caused by these units is related to the normalization of the matter power spectrum, which is commonly characterized in terms of the linear-theory rms mass fluctuation in spheres of radius 8h^-1 Mpc, σ8. This parameter does not correctly capture the impact of h on the amplitude of density fluctuations. We show that the use of σ8 has caused critical misconceptions for both the so-called σ8 tension regarding the consistency between low-redshift probes and cosmic microwave background data, and the way in which growth-rate estimates inferred from redshift-space distortions are commonly expressed. We propose to abandon the use of h^1 Mpc units in cosmology and to characterize the amplitude of the matter power spectrum in terms of σ12, defined as the mass fluctuation in spheres of radius 12Mpc, whose value is similar to the standard σ8 for h0.67.


CosmosClub: Erwan Allys (02/07/20)

CosmosClub Erwan Allys

Date: July 2nd 2020, 10.00 a.m.

Speaker: Erwan Allys (ENS Paris / École Normale Supérieure, Laboratoire de Radioastronomie )

Title:   The Wavelet Phase Harmonics, a new interpretable statistical description for analysis and synthesis of the LSS

Room: Zoom Meeting (connection details will be updated soon)


The statistical characterization of non-Gaussian fields is a major problem in current astrophysics, and no method has clearly emerged up to now to do so. In this presentation, I will introduce the Wavelet Phase Harmonics (WPH), a low-dimensional and interpretable set of statistics that efficiently characterizes the couplings between scales in non-linear processes. This description, that has been recently introduced in data science, is inspired from neural networks. Applied to projected matter density field from Quijote N-body Large Scale Structure (LSS) simulations, I will show how the WPH are able to provide better constraints on five cosmological parameters than the joint power spectrum and bispectrum, as well as to produce new realistic statistical syntheses from a maximum-entropy model. These results open the path to the use of a new type of statistical description for non-Gaussian fields in astrophysics.


CosmosClub: Florent Mertens (10/03/20)

Date: March 18th 2020, 10.30am

Speaker: Florent Mertens (LERMA / Kapteyn Astronomical Institute)

Title: The challenges of observing the Epoch of Reionization and Cosmic Dawn

Room: Cassini 


Low-frequency observations of the redshifted 21cm line promise to open a new window onto the first billion years of cosmic history, allowing us to directly study the astrophysical processes occurring during the Epoch of Reionization (EoR) and the Cosmic Dawn (CD). This exciting goal is challenged by the difficulty of extracting the feeble 21-cm signal buried under astrophysical foregrounds orders of magnitude brighter and contaminated by numerous instrumental systematics. Several experiments such as LOFAR, MWA, HERA, and NenuFAR are currently underway aiming at statistically detecting the 21-cm brightness temperature fluctuations from the EoR and CD. While no detection is yet in sight, considerable progress has been made recently. In this talk, I will review the many challenges faced by these difficult experiments and I will share the latest development of the LOFAR Epoch of Reionization and NenuFAR Cosmic Dawn key science projects.