Logo image
Back
Clustering of time series via non–parametric tail dependence estimation
Journal article   Peer reviewed

Clustering of time series via non–parametric tail dependence estimation

Fabrizio Durante, Roberta Pappadà and Nicola Torelli
Statistical Papers, Vol.56(3), pp.701-721
56
2015
DOI: https://doi.org/10.1007/s00362-014-0605-7
Handle:
https://hdl.handle.net/10863/1538

Abstract

Quantitative methods and economic modeling Clustering methods Time series Copulas Dependence models Quantitative risk management
We present a procedure for clustering time series according to their tail dependence behaviour as measured via a suitable copula-based tail coefficient, estimated in a non-parametric way. Simulation results about the proposed methodology together with an application to financial data are presented showing the usefulness of the proposed approach.
url
http://link.springer.com/article/10.1007/s00362-014-0605-7View

Details

Metrics

28 Record Views