Held regularly since 2001, the ADA – Astronomical Data Analysis – conference series has been focused on algorithms and information extraction from astrophysics data sets. It became a strongly interdisciplinary forum where researchers from diverse fields such as Astronomy, Statistics, Mathematics and Computing could interact. This conference series has been characterized by a range of innovative themes, including multi-scale geometric transforms such as the curvelet transform, compressed sensing and clustering, always remaining closely linked to front-line open problems and issues in Astronomy.

Along many of its editions, the ADA conference series included hands-on tutorial sessions on various topics of advanced data processing. In the era of large astronomical surveys that are grappling with unsolved methodological and data challenges, transforming Data into Science is a huge, and exciting, problem, and a large fraction of modern Astronomy depends on this. Nevertheless, postgraduate courses can hardly cope with the pace of development of the modern data analysis methods and tools that may enable researchers to make the best use of their data. This is the main focus of the ADA8 2016 Summer School: to present advanced data analysis methods and algorithms and to demonstrate how to use publicly available codes to improve knowledge extraction from astronomical datasets, enabling a better Science.

These tutorials are not intended exclusively for an audience with background in Astrononomy, and are aimed mainly aimed for young researchers at MSc, PhD and postdoc levels, albeit any researchers are welcome to attend. Moreover, this year ADA will take place on the days preceding the COSMO21 conference, both event take place at the same venue and the inscription in COSMO21 includes the possibility to participate in the ADA8 Summer School.

URL : http://ada8.cosmostat.org/

During the past 20 years there has been a resumption of a dialogue between astronomers and statisticians. This dialogue has been fruitful and has been the origin of a new discipline that is now widely called Astrostatistics. The main tools for comparing theoretical results with observations in astronomy are statistical. However, the development of huge astronomical databases presents challenges of scale, and has initiated an active use of newly-developed statistical techniques in astronomy, notable examples being sparsity and compressed sensing.

The meeting is especially timely from the point of view of cosmological surveys, where the size makes application of a fully Bayesian analysis computationally extremely demanding, especially in the realm of model selection. Pan-STARRS will have a complete survey of 3π steradians of petabyte size; the Dark Energy Survey and the VST KiDS surveys will be well underway presenting similar difficulties in the data analysis. Moreover, the cosmological community will be preparing for LSST and for Euclid, a survey of a large fraction of the sky at an angular resolution close to that of the Hubble Space Telescope. Wide-field spectroscopic cosmology surveys of will be targeting over 10 million objects with a spectral resolution of 5000, with the SKA precursors will be grappling with data challenges which currently are unsolved. These examples also highlight the big current role and even bigger future role of archival data in astrophysics research.

This meeting offers an opportunity to show-case techniques and methodologies that will have to be used by the wider community to use these archival data.

The fifth edition of the IN2P3 School Of Statistics gives an overview of the concepts and tools used in particle physics, astro-particle physics and cosmology when probabilities and statistics come to play. In particular, this edition is dedicated to the use of multivariate discriminants in the framework of data analyses performed in collider physics.
This school is targeted towards PhD students and towards senior physicists, aiming at extending their knowledge and skills in the field of statistical tools and frameworks developed for their fields.

The school takes place in Autrans at “l’Escandille” from Monday May 30th to Friday June 3rd 2016.
The lectures are subdivided into three parts: a part reminding the fundamental concepts used in Probabilities, Statistics and Hypothesis testing applied to physics analysis; a part focusing on the presentation of the concept and basis of most popular multivariate techniques and used in data analyses; and a part dedicated to actual multivariate tools and framework dedicated to data analyses in High energy Physics.
All lectures will be given in english. The School of Statistics is supported by the CNRS/IN2P3.

URL : https://indico.in2p3.fr/event/12667/

Annonce en lien avec l’Action/le Réseau : MAESTRO Formation

Thème :

Journées de l’Action MAESTRO

Présentation :

Dans le domaine de l’astronomie, l’évolution rapide des télescopes et autres instruments scientifiques ainsi que le recours intensif à la simulation informatique ont conduit, ces dernières années, à une production massive de données.

De plus l’Astrophysique doit aborder des problématiques de probabilités et statistiques dans le domaine de l’acquisition ou du traitement de données, de l’analyse et de l’exploitation (validation de modèles, problèmes inverses mal contraints, réduction de dimensionnalité, classifications, fouille de données, …).

Cette première réunion MAESTRO a pour but de mettre en contact les spécialistes d’astrophysique ainsi que les différents spécialistes en traitement de données, en stockage, mise à disposition et visualisation de données. Ces deux journées s’organiseront de manière à présenter les activités des laboratoires et équipes sur des réalisations astrophysiques et informatiques. Une partie des journées sera aussi dédiée à des discussions permettant un rapprochement des communautés.

Du : 2016-07-05

Au : 2016-07-06

Lieu : Salle des Thèses (à l’IRIT !)
Institut de Recherche en Informatique de Toulouse
Université Paul Sabatier UPS
118, route de Narbonne
F-31062 TOULOUSE Cedex

Site Web : https://maestro-2016.sciencesconf.org/

This school is addressed to doctorates and post-docs working in astrophysics or particle physics, with special emphasis on those working on LHCb, Belle II, Gaia, CTA or MAGIC.

The School will be structured around two Kaggle-challenge-like projects, one in the domain of particle physics and one in astrophysics which highlight the two big groups of Machine Learning and Data Mining techniques: classification and separation. With around 5-6 hours of lessons per day, the basics of these techniques will be covered, both from the theoretical and from the practical (hands-on) point of view. Two pieces of software, covering most needs, will be introduced and students will be free to use any of them for tackling the projects. Additional time will be devoted to covering some other innovative techniques.

URL : http://icc.ub.edu/congress/ICCUB_DM_SCHOOL/index.php

IAU symposium n.325 on Astroinformatics (AstroInfo16) will bring together world-class experts to address the methodological and technological challenges posed by the scientific exploitation of massive data sets produced by the new generation of telescopes and observatories. Astronomy, which already was at the forefront of Big Data science with exponentially growing data volumes and data rates, is now entering the petascale regime at optical, infrared and radio wavelengths.

The Symposium will cover a broad range of topics in astroinformatics: Database Management Systems, Data Mining, multiprocessor computing for astronomy, machine learning methods for classification and knowledge extraction, algorithms for N-point computations, time series analysis and image processing, advanced visualization for astronomical Big Data, cross-disciplinary perspectives and advanced training.

URL : http://dame.dsf.unina.it/astroinformatics2016.html

This is the third of a continuing series of annual workshops. Topics include: introduction to statistics; Bayesian inference, parameter estimation and model comparison, machine learning, density estimation, classification and dimensionality reduction. Lecturers are Roberto Trotta (Univ College London) and Zeljko Ivezic (Univ Washington).

URL : https://asaip.psu.edu/meetings/all-meetings/esac-data-analysis-and-statistics-workshop-2016

Astronomical observations produce some of the largest “big data” today through a new generation of telescopes. While these data sets exhibit the usual challenges associated with big data (immense data volumes, high dimensionality, high complexity, disparate variables, etc.) there are new problems such as pattern discovery from low-signal-to-noise imagery, near-real-time classification of multi-terabyte-size data, and the presence of strict laws of physics behind the data production which can often be assimilated into machine learning. This Special Session at a large IEEE Computational Intelligence meeting aims to engage the computational community in solutions to problems modern astronomy faces.

URL : https://asaip.psu.edu/organizations/ieee-astrominer-task-force/ieee-cidm-special-session-call-for-papers-mining-the-sky-knowledge-discovery-in-big-and-complex-astronomical-data-sets-and-data-streams

BigSkyEarth is organizing a Training School on “Visualization for large scale analytics” to be held at the University of Central Lancashire, UK, on April 3-8, 2017. A diverse set of key practitioners from the astronomy, the Earth observation and the computer science domains will be participating in the School, contributing the perspectives of both the academic and the industrial sector. Grants will be made available by the Action for a number of students.

URL : http://bigskyearth.eu/apply-to-the-bigskyearth-training-school-2017-visualization-for-large-scale-analytics/

Annonce en lien avec l’Action/le Réseau : MAESTRO

Thème :

Masse de données en Astrophysique

Présentation :

Dans le domaine de l’astronomie, l’évolution rapide des télescopes et autres instruments scientifiques ainsi que le recours intensif à la simulation informatique ont conduit, ces dernières années, à une production massive de données.

Dans le cadre des animations soutenues par le GdR MADICS, l’Action MAESTRO organise un atelier rassemblant des chercheurs, praticiens s’intéressant à la gestion des grandes masses de données en astrophysique. L’objectif de l’atelier est de faire échanger les différents participants sur les problématiques et/ou solutions envisagées afin de traiter efficacement les masses de données en jeu.

Nous sollicitons à cet effet des contributions pour des présentations sur les sujets suivants (liste non exhaustive).
* Retour d’expérience sur traitement de données en astrophysique
* Fouille interactive de grandes masses de données
* Optimisation de requêtes
* Parallélisme et données distribuées
* Analyse statistique et problèmes algorithmiques sous-jacents.
* Ontologies et données en astrophysique
* Intégration de données massives

Du : 2017-06-23

Au : 2017-06-23

Lieu : Ecole de Management Marseille

Montée de l’Université
Rue Joseph Biaggi – CS 70329
13331 Marseille Cedex 3

Site Web : https://maestro-2017.sciencesconf.org/

Programme :

• Introduction :
  • 10:00 MAESTRO (présentation), C. Surace/S. Maabout
  • 10:10 Plateforme Galactica, F. Gaudet
  • 10:30 Evolution of Data Management Systems for Big Data Applications, A. Hameurlain
• Données Astrophysiques :
  • 11:00 SKA, C. Ferrari
  • 11:30 LSST, E. Gangler
• 12:00 Déjeuner
• Machine Learning et Deep Learning
  • 13:30 Exploring the spectroscopic diversity of type Ia supernovae with DRACULA: a machine learning approach, E. Ishida
  • 13:45 Classification of reliability for redshift measurements, S. Jamal
  • 14:00 Clustering pour détection d’amas ouverts avec Gaia, M. Morvan
  • 14:15 Galaxy morphology with CNNs using transfer learning, A. Boucaud
• Organisations :
  • 14:30 BIGSKYEARTH (présentation), E. Gangler/K. Zeitouni
  • 14:50 Cosmostatistics initiative : (https://asaip.psu.edu/organizations/iaa/iaa-working-group-of-cosmostatistics), E. Ishida
• 15:10 Wrap up, sondage, conclusions

Programme
Finding hidden correlations in the complex astronomical Big Data
Challenges of spatial queries in peta-scale surveys
Massively parallel data mining in huge distributed databases
Scientific visualisation of complex many-dimensional data sets
Standardization of meta-data for better science
Advanced statistical inference in cosmology
Real time transient detection and classification
Applications of machine learning for source classification and clustering
Deep neural networks for feature extraction and image classification
Active learning and Domain Adaptation
Interdisciplinary panel discussions

URL : http://eas.unige.ch/EWASS/session.jsp?year=2017&id=S14

Annonce en lien avec l’Action/le Réseau : MAESTRO /

Thème :

Astrostatistique

Présentation :

L’objectif de l’école est de donner les compétences nécessaires aux participants pour entreprendre par eux-mêmes des analyses de type bayésien qui deviennent de plus en plus répandues en astrophysique (cosmologie, exoplanètes…). Nous insistons également beaucoup sur l’importance d’établir des collaborations entre astrophysiciens et statisticiens pour des développements algorithmiques et méthodologiques souvent requis par la spécificité des données astrophysiques.

Du : 2017-10-09

Au : 2017-10-13

Lieu : L’ESCANDILLE VILLAGE VACANCES – Autrans (Vercors)

Site Web : https://stat4astro2017.sciencesconf.org/

Annonce en lien avec l’Action/le Réseau : MAESTRO /

Thème :

Astrophysique, machine learning, big data

Présentation :

Objectifs de la formation

– Sensibiliser les chercheurs et ingénieurs aux nouvelles technologies du traitement de données,
– Former les scientifiques et ingénieurs aux méthodes informatiques de traitement de masse données en Astrophysique,
– Démontrer les possibilités d’utiliser ces nouvelles méthodes dans le cas de données astrophysiques à travers des cas pratiques,
– Sensibiliser et Former les personnels au traitement de grande masse de données,
– Optimiser les utilisations de plateformes de traitement telle que GALACTICA.

Un Hackaton est prévu en fin de semaine afin d’initier des projets de collaboration (venez avec vos projets)

Du : 2018-06-25

Au : 2018-06-29

Lieu : Polytech Marseille – Parc scientifique et technologique de Luminy

163 avenue de Luminy – 13288 Marseille Cedex 09

Site Web : https://astroinfo2018.sciencesconf.org/

Annonce en lien avec l’Action/le Réseau : MAESTRO

Thème :

Traitement d’image en astrophysique

Présentation :

Nous organisons les 24 et 25 janvier 2019 à Strasbourg deux journées scientifiques sur le traitement d’images astronomiques. Ces deux journées seront ouvertes aux chercheurs en traitement d’images ainsi qu’aux astronomes, l’idée principale étant de stimuler les discussions entre ces deux disciplines sans se restreindre à une problématique ou un instrument. L’objectif de ces journées est double. Tout d’abord, elles permettront de faire le point sur les recherches actuelles et les problèmes ouverts, qu’ils soient actuels ou pressentis. Ensuite, elles favoriseront les échanges entre les deux communautés, de façon à réfléchir aux verrous existants et à préparer la résolution des problèmes ouverts.

Du : 2019-01-24

Au : 2019-01-25

Lieu : Strasbourg

Site Web : https://jatia2019.sciencesconf.org/

Annonce en lien avec l’Action/le Réseau : MAESTRO

Thème :

Astrostatistics

Présentation :

The topic of the 2019 session of Stat4Astro is the time series (including variabilities and transient events) that, from celestial mechanics to gravitational waves, from exoplanets to quasars, concern nearly all the astrophysics. Variable phenomena are ubiquitous in the Universe: periodic (orbits, cycles, pulses, rotations…), transient (explosions, bursts, stellar activity…), random (accretion, ejection…) or regular (apparent motions…). The detection, the characterization and the classification of these variabilities is a discipline of statistics called time series analysis. In astrophysics, the detection can be immediate to alert other telescopes, or very detailed to identify some exoplanets or probe the interior of stars. The characterization is required for the physical modeling and understanding. Classification is of course necessary to organize the observations.

Du : 2019-10-06

Au : 2019-10-11

Lieu : Autrans (France)

Site Web : https://stat4astro2019.sciencesconf.org