Practical applications of Machine Learning in modern Astronomy

Europe/Berlin
ZOOM (DESY)

ZOOM

DESY

Platanenallee 6 15738 Zeuthen
Description
During this interactive, online workshop, you will learn about common concepts and tools in modern ML, and work through examples of their application to problems in astronomical data analysis.
Participants
  • Amy Miller
  • Andrea Porelli
  • Anke Arentsen
  • Artem Bohdan
  • Bastian-Querner Benjamin
  • Christian Dersch
  • Christine Greif geb. Köpferl
  • Christoph Kuckein
  • Cristina Lagunas Gualda
  • Dalal El Youssoufi
  • Ekaterina Dineva
  • Erasmo Trentin
  • Federica Bradascio
  • Genoveva Micheva
  • Helge Todt
  • Heshou Zhang
  • Huirong Yan
  • Jannis Necker
  • Jean Damascene Mbarubucyeye
  • Jennifer Wojno
  • Jesper Storm
  • Julian Schliwinski
  • Kris Youakim
  • Martin Wendt
  • Meetu Verma
  • Nikolay Kacharov
  • Pedro Ivo Silva Batista
  • Raul Ribeiro Prado
  • Robert Stein
  • Samir Nepal
  • Simeon Reusch
  • Simone Garrappa
  • Taavi Tuvikene
  • Thomas Schmidt
  • Victor Barbosa Martins
  • Tuesday 17 November
    • 09:00 09:45
      Introduction to Machine Learning: different types of machine learning algorithms, Theoretical background, practical aspects, structure and elements of (convolutional) neural networks 45m
      Speaker: Gal Matijevič
    • 09:45 10:30
      Unsupervised Learning I: application of t-SNE and clustering to big datasets 45m
    • 10:30 11:00
      Break 30m
    • 11:00 12:30
      Supervised Learning: practical introduction to convolutional neural networks and application to stellar magnitude regression (Part 1) 1h 30m
    • 12:30 13:30
      Break 1h
    • 13:30 15:00
      Supervised Learning: practical introduction to convolutional neural networks and application to stellar magnitude regression (Part 2) 1h 30m
  • Wednesday 18 November
    • 09:00 10:30
      Unsupervised Learning II: autoencoders used on stellar spectra 1h 30m
    • 10:30 11:00
      Break 30m
    • 11:00 12:30
      Supervised Learning II: solar image segmentation with convolutional neural networks 1h 30m
    • 12:30 13:30
      Break 1h
    • 13:30 15:00
      Supervised Learning III: Generative Adversarial Networks (GANs) used to generate realistic galaxy images 1h 30m