Forecasting and granger modelling with non-linear dynamical dependencies

Gregorovà, Magda (Haute école de gestion de Genève, HES-SO // Haute Ecole Spécialisée de Suisse Occidentale ; University of Geneva, Switzerland) ; Kalousis, Alexandros (Haute école de gestion de Genève, HES-SO // Haute Ecole Spécialisée de Suisse Occidentale ; University of Geneva, Switzerland) ; Marchand-Maillet, Stéphane (Haute école de gestion de Genève, HES-SO // Haute Ecole Spécialisée de Suisse Occidentale)

Traditional linear methods for forecasting multivariate time series are not able to satisfactorily model the non-linear dependencies that may exist in non-Gaussian series. We build on the theory of learning vector-valued functions in the reproducing kernel Hilbert space and develop a method for learning prediction functions that accommodate such non-linearities. The method not only learns the predictive function but also the matrix-valued kernel underlying the function search space directly from the data. Our approach is based on learning multiple matrix-valued kernels, each of those composed of a set of input kernels and a set of output kernels learned in the cone of positive semi-de_nite matrices. In addition to superior predictive performance in the presence of strong non-linearities, our method also recovers the hidden dynamic relationships between the series and thus is a new alternative to existing graphical Granger techniques.


Type de conférence:
full paper
Faculté:
Economie et Services
Ecole:
HEG - Genève
Institut:
CRAG - Centre de Recherche Appliquée en Gestion
Classification:
Economie/gestion
Adresse bibliogr.:
Skopje, Macedonia, 18-22 September 2017
Date:
Skopje, Macedonia
18-22 September 2017
2017
Pagination:
16 p.
Publié dans
Proceedings of the European Conference on Machine Learning & Principles and Practice of Knowledge Discovery in Databases
Ressource(s) externe(s):
Le document apparaît dans:



 Notice créée le 2017-10-18, modifiée le 2018-12-07

Fichiers:
Télécharger le document
PDF

Évaluer ce document:

Rate this document:
1
2
3
 
(Pas encore évalué)