Skip to Main content Skip to Navigation

Un algorithme de découverte de chroniques pertinentes pour le diagnostic par identification et reconstitution

Alexandre Sahuguède 1
1 LAAS-DISCO - Équipe DIagnostic, Supervision et COnduite
LAAS - Laboratoire d'analyse et d'architecture des systèmes
Abstract : Chronicles are temporal patterns well-suited for an abstract representation of complex dynamic systems. Chronicle recognition algorithms allow the identification of chronicles in an on-line stream of data to be done and take adequate action in an quick and efficient manner. Chronicles are used in a vast array of applications such as medical field, internet networks, or industrial applications. Nevertheless, designing chronicles is not an easy task due to the sophistication and the increase of data generation capacity of modern systems. The chronicle discovery process try and tackle this problem by an automatic design of chronicles from data directly generated by the studied system. In this thesis, an innovative approach to the problem of chronicle discovery is introduced. This new approach lies of the identification of elementary chronicles and a reconstitution of complex chronicles from them. The algorithm introduced, called CDIRe (Chronicle Discovery by Identification and Reconstitution), allows the discovery of chronicles with few knowledge from the underlying system to be done.
Complete list of metadata
Contributor : Abes Star :  Contact
Submitted on : Monday, March 8, 2021 - 5:28:58 PM
Last modification on : Thursday, June 10, 2021 - 3:06:26 AM


Version validated by the jury (STAR)


  • HAL Id : tel-02880629, version 2


Alexandre Sahuguède. Un algorithme de découverte de chroniques pertinentes pour le diagnostic par identification et reconstitution. Algorithme et structure de données [cs.DS]. Université Paul Sabatier - Toulouse III, 2020. Français. ⟨NNT : 2020TOU30166⟩. ⟨tel-02880629v2⟩



Record views


Files downloads