Day-ahead time series forecasting: application to capacity planning

Abstract : In the context of capacity planning, forecasting the evolution of informatics servers usage enables companies to better manage their computational resources. We address this problem by collecting key indicator time series and propose to forecast their evolution a day-ahead. Our method assumes that data is structured by a daily seasonality, but also that there is typical evolution of indicators within a day. Then, it uses the combination of a clustering algorithm and Markov Models to produce day-ahead forecasts. Our experiments on real datasets show that the data satisfies our assumption and that, in the case study, our method outperforms classical approaches (AR, Holt-Winters).
Type de document :
Communication dans un congrès
AALTD'18 - 3nd ECML/PKDD Workshop on Advanced Analytics and Learning on Temporal Data, Sep 2018, Dublin, Ireland
Liste complète des métadonnées

https://hal.inria.fr/hal-01912002
Contributeur : Colin Leverger <>
Soumis le : lundi 5 novembre 2018 - 08:20:04
Dernière modification le : mardi 12 février 2019 - 10:15:56
Document(s) archivé(s) le : mercredi 6 février 2019 - 13:23:01

Fichiers

ts_workshop_cl.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01912002, version 1
  • ARXIV : 1811.02215

Citation

Colin Leverger, Vincent Lemaire, Simon Malinowski, Thomas Guyet, Laurence Rozé. Day-ahead time series forecasting: application to capacity planning. AALTD'18 - 3nd ECML/PKDD Workshop on Advanced Analytics and Learning on Temporal Data, Sep 2018, Dublin, Ireland. 〈hal-01912002〉

Partager

Métriques

Consultations de la notice

117

Téléchargements de fichiers

13