Title: A clustering procedure for studying financial integration with big data time series
Authors: Pilar Poncela - Universidad Autonoma de Madrid (Spain)
Nuno Crato - EC-JRC (Italy) [presenting]
Jorge Caiado - Technical University of Lisbon (Portugal)
Abstract: Time and frequency-domain procedures for characterizing and comparing large sets of long time series are studied. The procedures are computationally simple, condense second-order information of the time series under consideration, and develop similarity and dissimilarity measures for comparing these time series. On the basis of these measures, we exemplify and compare various ways of clustering time series. Instead of using all the information available from data, which would be computationally very expensive, we propose some regularization rules in order to select and summarize the most relevant information for clustering purposes. We use these procedures to study the evolution of several stock markets indices, extracting information on the European financial integration. We further show the effect of the recent financial crisis over these indices behavior.