Summer School

An introduction to the Summer School


The Institute of Space Sciences (ICE-CSIC) does forefront scientific and technological research with the mission of contributing to our understanding of the Cosmos.

Our centre organises a Summer School for Master and PhD students every year over the summer, usually during the month of July.

Students from everywhere have the opportunity to broaden their knowledge, network with researchers and use the experience acquired for their thesis, PhD or post-doc research in the future.

The theme of the 4th Institute of Space Sciences Summer School (2021) was Artificial Intelligence for Astronomy.

4th Summer School

Artificial Intelligence
for Astronomy

Astronomy and astrophysics are undergoing a rapid evolution towards data-driven science. Large-scale astronomical surveys and computer simulations are creating enormous amounts of datasets that require the development of new techniques for their analysis. Both complexity and sheer size of these datasets, and often the necessity of combining heterogenous sources of data, put machine learning methods as a central tool, present and future, for scientific discovery in astronomy.

The program of the 4th Institute of Space Science Summer School will focus on artificial intelligence (AI) methods for astronomy research, with special focus on neural networks for image classification, natural language processing and graph neural networks. The topics will cover the mathematical concepts as well as the development of software tools and applications. There will be lectures and hands-on activities.

Program

Lectures

  • Machine Learning Overview
  • Introduction to Deep Learning
  • Convolutional Neural Networks
  • Fully connected Neural Networks
  • Advance Deep Learning
  • Unsupervised Representation
  • Natural Language Processing
  • Graph Neural Networks
  • Generative models
  • Applications in astronomy
  • Applications outside astronomy

Hands-on sessions

  • Introduction to ML
  • Neural network and backpropagation from scratch with Numpy
  • Image classification

 

© ElenaBor

Where?

Every year, Summer School takes place at the Institute of Space Sciences building, located at Universitat Autónoma de Barcelona (UAB) campus. In 2021, due to the pandemic, it took place online.

Deadlines

The Institute of Space Sciences welcomed around 40 Master and PhD students to the Summer School, in which they broadened their knowledge on this exciting field as well as got into contact with research groups working at the Institute in this and other fields. Registration deadlines for 2021 are listed below.

◉ Registration opened from 20/05/2021 until 20/06/2021

◉ Selected participants were notified by 30/06/2021

◉ School ran from 12/07/2021 to 16/07/2021

Lecturers 2021

Meet the host lecturers of the 4th Summer School. 

  • Helena Domínguez

  • Vanessa Graber

  • Alessandro Patruno

    Guest Lecturers

    Stefano Bocconi - Zephyros Solutions (The Netherlands)

    Stefano Bocconi has worked as a researcher in computer science in the Netherlands at TU Delft, VU University Amsterdam, and Elsevier and in Italy at the University of Turin. He obtained his PhD “Automatic Generation of Video Documentaries" at CWI, Amsterdam in 2005, and has a master degree in Electrical Engineering from the University of Florence, Italy. He now collaborates as a software developer for the startup Zephyros Solutions.

    Marc Huertas Company - Instituto Astrofísica Canarias (Spain)

    Marc Huertas-Company is an associate professor at the Paris Observatory and the University of Paris, on leave from the Instituto de Astrofísica de Canarias with a Ramon y Cajal Fellowship. He is also a member of the prestigious Institut Universitaire de France. He studies how galaxies form and evolve using artificial intelligence to link observations with current theoretical knowledge of galaxy formation. He also teaches Deep Learning techniques to grad students in astrophysics and at several international schools.

    Alison Lowndes - NVIDIA (UK)

    Alison Lowndes is dedicated to providing solutions with AI & deep learning, to some of humanity's biggest problems. Working directly with entities like NASA, SETI, ESA and the United Nations, thanks to NVIDIA, she looks over the global research field, staying up to date with progress from Luminaries to Masters students and talking to everyone about how we enable AI. She is also founding team member of the NASA Frontier Development Lab, an AI R&D accelerator that tackles knowledge gaps useful to the Space program and studies on topics not only important to NASA, but also to Humanity’s future.

    Andrey Malakhov - Zephyros Solutions (The Netherlands)

    Andrey Malakhov is an expert in machine learning and a software developer with a broad and diverse track-record. He has worked for several companies and institutions as a data scientist, software developer and modeller. His background is in econometrics and finance with a M.Sc. in finance obtained in 2011 from the Vrije Universiteit in Amsterdam. He has co-founded the startup Zephyros Solutions and has worked on several deep learning projects with a social impact.

    Pau Rodríguez López - Element AI (Canada)

    Pau Rogríguez López is a research scientist at ElementAI Montreal. His research interests cover computer vision, unsupervised learning, meta-learning, and generalization. He did his Ph.D. at the Computer Vision Center (UAB) on deep learning for fine-grained image recognition. Previously, he obtained an MSc in Artificial Intelligence from KU Leuven (Belgium). Currently, he is working on unsupervised learning algorithms for satellite imaging with minimal human annotation. He is convinced that machine learning and space sciences have great potential of cross-fertilization.

     

    Past editions

    2020

    Suspended due to the Covid-19 health crisis

    2019

    Geoscience and Remote Sensing society (GRSS)

    2018

    Gravitational Wave Astronomy

    2017

    Neutron Stars and their Environments


    © 2021 Institute of Space Sciences (ICE-CSIC). All rights reserved.