13–17 May 2019
ESTEC, Noordwijk, The Netherlands
Europe/Zurich timezone

Session

Artificial intelligence highlight

16 May 2019, 14:00
ESTEC

ESTEC

Presentation materials

There are no materials yet.

  1. Jose Martinez-Heras (ESA)
    16/05/2019, 14:00

    We will talk about Artificial Intelligence and Machine Learning (ML). In this introduction we will talk about what machine learning is, how it works, types of machine learning, some example applications in industry & space. We will discuss the challenging aspect of running successful machine learning projects such as the ML workflow and how to make ML generalise better. We will touch upon how...

    Go to contribution page
  2. Lorenzo Moneta (CERN)
    16/05/2019, 14:45

    We will focus the lecture on artificial neural networks and deep learning. We will give a theoretical introduction to these machine learning methods presenting commonly
    used architectures such as convolutional neural networks, recurrent network and generative adversarial networks.
    In addition, we will provide examples how deep learning methods are currently used in High Energy physics...

    Go to contribution page
  3. Jose Martinez-Heras (ESA)
    16/05/2019, 15:45

    In this hands-on session we will use machine learning to predict the thermal power consumption of Mars Express. A good prediction is needed in order to make use of as many scientific instruments on-board at possible. Since we talked about decision trees and random forests during the theoretical part, we will use them during this practical session in python.

    Go to contribution page
  4. Lorenzo Moneta (CERN)
    16/05/2019, 17:00

    In the hands-on session we will present machine learning software such as Keras to build and train deep learning model
    We will show an example of event classification using open simulated data of an LHC experiment and an example for building a generative adversarial network (GAN) to generate images.
    The hands-on session will be run in Python using the CERN SWAN service.

    Go to contribution page
Building timetable...