Third ML-INFN Hackathon: Advanced Level

Europe/Rome
Sala Multimediale

Sala Multimediale

INFN Sezione di Bari Via Giovanni Amendola, 173, 70126 Bari BA
Alfonso Monaco (BA), Domenico Diacono (Istituto Nazionale di Fisica Nucleare), Lucio Anderlini (Istituto Nazionale di Fisica Nucleare), Nicola De Filippis (BA), Tommaso Boccali (Istituto Nazionale di Fisica Nucleare)
Description

Welcome to the third edition of the Machine Learning @INFN (ML_INFN) advanced level hackathon.  This edition is hosted at Dipartimento Di Fisica, INFN Sezione di Bari.

Notably, it is the first Hackathon to happen in Person, so please apply only if you are planning to come to Bari. The logistics allow for ~ 20 participants.

You can check the first edition here and the second edition here, with most of the talks attached as video files.

The mandatory registration process will be open from Monday October 12th to Monday October 24th. 

In case of a number of registrations exceeding the available positions, the applications will be ranked and selected on the basis of the scientific CV of the applicants.

The successful applicant will be informed by November 1st. Please do not book hotel/flight before a positive confirmation.

The course is to be considered as "advanced level" for Machine Learning topics. The hackathon will be organized over 4 days, distributed as 

  1. Infrastructure and tools for Machine Learning
  2. Models constitued of multiple neural networks
  3. Models for data beyond tabular format
  4. Ongoing developments towards the future of Machine Learning

 The afternoons of days 2 and 3 will be devoted to experimenting the various methods and architectures with the help of tutors.

Upon registration, users will be asked to express their preferences for a one of the use cases offered. We will try to

  • whenever possible, satisfy the preference in the order given
  • try to form groups with students with the full range of proficiencies, in order to allow for self-tutoring inside the groups

The list of available use cases for the hackathon are currently (there could be additions depending on the registration process and on the status of other opportunities):

  1. Lung segmentation on Chest X-Ray images with U-Net
  2. Domain adaptation to train model-indepentent classifiers in High Energy Physics
  3. Graph Neural Networks and Transformers
  4. Explainable Artificial Intelligence (XAI)

More information is available in the Hackathon exercises tab on left menu.

 

Prerequisites

Technical: access to INFN Cloud provided resources

NOTE: users will use INFN-Cloud resources and this does not require any specific INFN-Cloud authorization. 

Please read and comply with the recommendations in the side menu "technical prerequisite".

Machine learning related knowledge

The hackathon is to be considered as advanced level for Machine Learning. Hence, students are expected to have a starting level of understanding of machine learming and on the technologies it implies. For exemple

  • good fluency in Python:
    • program flow, definition and use of variables
    • definition and utilization of functions
    • package management (import, install via pip)
  • fluency in numpy:
    • arrays, basic operations, reshaping
  • fluency on matplotlib:
    • how to define and plot a basic figure
  • fluency on online python notebooks
    • (like SWAN, Jupyter, Google Colab)
  • basic concepts of Deep Learning
    • (e.g.) working principle of dense and convolutional layers 
    • epochs, batches, optimizers, metrics...

Registration

Attending the hackathon is free and happens via the entry in the side menu.

The number of participants can be limited, depending on the tutoring and technical capabilities.

Upon registration, we will require:

  • INFN Affiliation
  • Research interests
  • Proficiency level in Machine Learning and related tools
  • Specific interests in Machine learning  techniques
  • The preferred use case for the third day (first and second choice)

For INFN employees

The event was approved by the INFN National Committee for Continuos Education (Commissione Nazionale Formazione). INFN Employees participating to the event can cover travel expenses with the funds labelded "Formazione" of their own INFN units. 

INFN Employees taking part to the event are expected to notify the local responsible for Continuos Education of their unit to receive the mandatory endorsement from the unit's Director.

At the end of the event, the list of the partecipating INFN employees will be recorded in the national database of continuos education activities. 

Additional Material

Please consider visiting again this page from time to time: we will add more resources for the hackathon and related projects as soon as they become available,.

Participants
  • Alessandro Bombini
  • Alkis Papanastassiou
  • Andrea Di Luca
  • Antonio Sidoti
  • Benedetta Camaiani
  • Biswarup Paul
  • Domenico Pomarico
  • Enrico Calore
  • Fabio Catalano
  • Fernando Garcia-Avello Bofias
  • Filippo Santoliquido
  • Francesco Maria Follega
  • Giacomo Serra
  • Gianfranco Paternò
  • Giovanni Francesco Tassielli
  • Giuseppe Di Somma
  • Lorenzo Marini
  • Luca Giommi
  • Marco Serra
  • Matteo Lai
  • Paolo Mastrandrea
  • Riccardo Scheda
  • Sara Saponaro
    • 2:25 PM 6:20 PM
      Cloud Infrastructure for Interactive Analysis and Machine Learning: Monday
      Convener: Tommaso Boccali (Istituto Nazionale di Fisica Nucleare)
      • 2:25 PM
        Welcome by the local authorities: Departiment Chair and INFN Bari director 5m
      • 2:30 PM
        Welcome and introduction 10m
        Speaker: Tommaso Boccali (Istituto Nazionale di Fisica Nucleare)
      • 2:45 PM
        INFN Cloud and National Services 20m
        Speaker: Giacinto Donvito (Istituto Nazionale di Fisica Nucleare)
      • 3:10 PM
        HPC resources at CNAF and CINECA 20m
        Speaker: Stefano Dal Pra (Istituto Nazionale di Fisica Nucleare)
      • 3:35 PM
        ReCaS-Bari, datacenter and HPC resources 20m
        Speaker: Domenico Di Bari (Istituto Nazionale di Fisica Nucleare)
      • 4:00 PM
        Coffee break 25m
      • 4:30 PM
        Open Science Cloud 25m
        Speaker: Daniele Spiga (Istituto Nazionale di Fisica Nucleare)
      • 5:00 PM
        Hands-on ML_INFN instances, with Jupyter and beyond 1h
        Speaker: Lucio Anderlini (Istituto Nazionale di Fisica Nucleare)
    • 9:00 AM 1:00 PM
      Advanced Machine Learning Lectures
    • 1:00 PM 2:30 PM
      Lunch Break 1h 30m
    • 2:30 PM 6:30 PM
      Hackathon exercises: Tuesday
      • 2:30 PM
        Hackathon: Lung Sementation on Chest X-Ray images with U-Net 2h Sala riunioni ReCaS (Centro ReCaS)

        Sala riunioni ReCaS

        Centro ReCaS

        Speakers: Francesca Lizzi (Istituto Nazionale di Fisica Nucleare), Ian Postuma (Istituto Nazionale di Fisica Nucleare)
      • 2:30 PM
        Hackathon: Lung Sementation on Chest X-Ray images with U-Net 2h Sala riunioni ReCaS (Centro ReCaS)

        Sala riunioni ReCaS

        Centro ReCaS

        Speakers: Francesca Lizzi (Istituto Nazionale di Fisica Nucleare), Ian Postuma (Istituto Nazionale di Fisica Nucleare)
      • 4:30 PM
        Coffee break 30m Sala Multimediale

        Sala Multimediale

        INFN Sezione di Bari Via Giovanni Amendola, 173, 70126 Bari BA
      • 5:00 PM
        Hackathon: Domain Adaptation in HEP 1h Sala Giunta (Dipartimento di Fisica)

        Sala Giunta

        Dipartimento di Fisica

        Speakers: Lorenzo Viliani (Istituto Nazionale di Fisica Nucleare), Piergiulio Lenzi (Istituto Nazionale di Fisica Nucleare)
      • 5:00 PM
        Hackathon: Lung Sementation on Chest X-Ray images with U-Net 1h Sala riunioni ReCaS (Centro ReCaS)

        Sala riunioni ReCaS

        Centro ReCaS

        Speakers: Francesca Lizzi (Istituto Nazionale di Fisica Nucleare), Ian Postuma (Istituto Nazionale di Fisica Nucleare)
    • 8:15 PM 10:15 PM
      Social Dinner 2h
    • 9:00 AM 1:00 PM
      Advanced Machine Learning Lectures
    • 1:00 PM 2:30 PM
      Lunch break 1h 30m
    • 2:30 PM 6:30 PM
      Hackathon exercises: Wednesday
      • 2:30 PM
        Explainable AI on genetic data 2h Sala Giunta (Dipartimento di Fisica)

        Sala Giunta

        Dipartimento di Fisica

        Speakers: Alfonso Monaco (BA), Domenico Diacono (Istituto Nazionale di Fisica Nucleare), Sabina Tangaro (Istituto Nazionale di Fisica Nucleare)
      • 2:30 PM
        Hackathon: GNN e Transformers 2h Sala Multimediale

        Sala Multimediale

        INFN Sezione di Bari Via Giovanni Amendola, 173, 70126 Bari BA
        Speakers: Angela Taliercio, Claudio Caputo (BA), Stefano Giagu (Istituto Nazionale di Fisica Nucleare)
      • 4:30 PM
        Coffee break 30m Sala Multimediale

        Sala Multimediale

        INFN Sezione di Bari Via Giovanni Amendola, 173, 70126 Bari BA
      • 5:00 PM
        Explainable AI on genetic data 1h Sala Giunta (Dipartimento di Fisica)

        Sala Giunta

        Dipartimento di Fisica

        Speakers: Alfonso Monaco (BA), Domenico Diacono (Istituto Nazionale di Fisica Nucleare), Sabina Tangaro (Istituto Nazionale di Fisica Nucleare)
      • 5:00 PM
        Hackathon: GNN e Transformers 1h Sala Multimediale

        Sala Multimediale

        INFN Sezione di Bari Via Giovanni Amendola, 173, 70126 Bari BA
        Speakers: Angela Taliercio, Claudio Caputo (BA), Stefano Giagu (Istituto Nazionale di Fisica Nucleare)
    • 9:00 AM 9:30 PM
      Ongoing developments and future scenarios
      • 9:00 AM
        Bayesian hyperparameter optimization 55m
        Speaker: Matteo Barbetti (INFN Firenze)
      • 10:00 AM
        Machine Learning on FPGA: Introduction 25m
        Speaker: Mirko Mariotti (Istituto Nazionale di Fisica Nucleare)
      • 10:30 AM
        Coffee break 25m
      • 11:00 AM
        Machine Learning on FPGAs: Hands-on BondMachine 1h 25m
        Speaker: Mirko Mariotti (Istituto Nazionale di Fisica Nucleare)
      • 12:30 PM
        Closing remarks and outlook 25m
        Speakers: Lucio Anderlini (Istituto Nazionale di Fisica Nucleare), Tommaso Boccali (Istituto Nazionale di Fisica Nucleare)