A process mining approach to the analysis of the structure of time series

Par Conseil national de recherches du Canada

Télécharger
  1. (PDF, 1 Mo)
DOITrouver le DOI : https://doi.org/10.1007/978-3-030-63089-8_25
AuteurRechercher : 1; Rechercher : ; Rechercher : 2; Rechercher :
Affiliation du nom
  1. Conseil national de recherches du Canada. Technologies numériques
  2. Conseil national de recherches du Canada. Herzberg en astronomie et en astrophysique
FormatTexte, Chapitre de livre
Sujetprocess mining; machine learning; graph an trace clustering; fuzzy models; evolutionary multi-objective models
Résumé
This paper presents a discussion of the potential of Process Mining for the analysis of general processes involving the time variation of magnitudes given by real-valued variables. These scenarios are common in a broad variety of domains, like natural and life sciences, engineering, and many others beyond business processes, where, in general, complex systems are observed and monitored using sensor data, producing time-series information. Two approaches are presented to construct event logs for such types of problems and one of them is applied to a real-world case (monitoring the F10.7 cm electromagnetic flux produced by the Sun). The results obtained with the Fuzzy Miner and the Multi-Objective Evolutionary Tree Miner algorithms successfully exposed the differences in the internal structure of the F10.7 cm series between Solar cycles. For this application, Process Mining proved to be a valuable tool for analyzing the rhythm of solar activity and how it is changing. The approach introduces here is general and could be used in the analysis of data from a broad variety of time-dependent information from many domains.
Date de publication
Date de création2020-11-30
Maison d’éditionSpringer
Dans
  • Proceedings of the Future Technologies Conference (FTC) 2020, Volume 2 : 379–392.
Série
  • Advances in Intelligent Systems and Computing 1289.
Langueanglais
Publications évaluées par des pairsOui
Exporter la noticeExporter en format RIS
Signaler une correctionSignaler une correction
Identificateur de l’enregistrement292c8b72-091a-467a-8be3-bb3fe0499bc7
Enregistrement créé2020-11-30
Enregistrement modifié2020-11-30
Signaler un problème sur cette page
Date de modification: