STAGE Design and implementation of an Change Point Detection web application

Date de mise à jour de l’offre

ECOLE NORMALE SUPERIEURE PARIS-SACLAY :

L'École normale supérieure Paris-Saclay est un établissement d’enseignement supérieur et de recherche (de L3 jusqu'au doctorat), qui fait partie de l'Université Paris-Saclay.

Description de la mission

A common task in signal processing is the identification and analysis of complex systems whose underlying state changes, possibly several times. This setting arises when industrial systems, physical phenomena or human activity are continuously monitored with sensors. The objective of practitioners is to extract from the recorded signals a posteriori meaningful information about the different states and transitions of the monitored object for analysis purposes. This setting encompasses a broad range of real-world scenarios and a wide variety of signals. Change point detection is the task of finding changes in the underlying model of a signal or time series [1].

Change point detection methods rely on several parameters such as the cost function (which determines the type of change-points that will be detected) or the penalty (which is linked to the sensitivity of the detector). Setting these parameters can be fastidious and remains the main obstacle to the diffusion of such methods into application domains such as healthcare. Recently, Centre Borelli has proposed several contributions to this problem: first, the diffusion of an open-source Python library (ruptures [2]) which has been downloaded more than 1 million times, and second, two preliminary algorithms for automatically learning the cost function [3] and the penalty [4] from annotated data.

The aim of this internship is to improve and integrate the two supervised approaches described in [3] and [4] and to build a prototype interactive web application.

The online application will allow users to
1. upload data,
2. annotate them (either by specifying change-points or homogeneous results)
3. learn the parameters of the algorithms and display the results
4. interactively play with the results to improve parameter learning.

This internship will lead to a publication and high visibility since the online demo will be diffused in several scientific communities.

References
[1] C. Truong, L. Oudre, N. Vayatis. Selective review of offline change-point detection methods. Signal Processing, 167:107299, 2020
[2] https://centre-borelli.github.io/ruptures-docs/
[3] C. Truong, L. Oudre and N. Vayatis. Supervised kernel change point detection with partial annotations. In Proceedings of the International Conference on Acoustics, Speech, and Signal Processing (ICASSP), pages 3147-3151, Brighton, UK, 2019.
[4] C. Truong, L. Oudre and N. Vayatis. Penalty Learning for Change-point Detection. In Proceedings of EUSIPCO.

Profil recherché

The candidate must have a particular interest in the development of interactive web applications (with a special emphasis on interactive annotation), and willing to learn and understand about the underlying algorithms. The fundamental part is building the interactive web application. The intern is expected to be able to develop with JQuery, SQL, Python 3, and GNU/Linux servers. Django is a bonus.

Niveau de qualification requis

Bac + 4/5 et +
  • Employeur
    ECOLE NORMALE SUPERIEURE PARIS-SACLAY
  • Secteur d’activité de la structure
    Enseignement - Formation - Recherche
  • Effectif de la structure
    Plus de 250 salariés
  • Site internet de la structure
    https://ens-paris-saclay.fr
  • Type de stage ou contrat
    Contrat de professionnalisation
  • Date prévisionnelle de démarrage
  • Durée du stage ou contrat
    Plus de 4 mois et jusqu'à 6 mois
  • Le stage est-il rémunéré ?
    Oui
  • Niveau de qualification requis

    Bac + 4/5 et +
  • Lieu du stage
    École normale supérieure Paris-Saclay Centre Borelli UMR 9010
    4, avenue des Sciences
    91190 GIF SUR YVETTE
  • Accès et transports
    https://ens-paris-saclay.fr/lecole/venir-lecole