Events and Meetings of Italian Statistical Society, Advances in Latent Variables - Methods, Models and Applications

Font Size: 
Dimension reduction in the time domain: Dynamic Principal Components
Daniel Peña

Last modified: 2013-06-17

Abstract


This talk presents Dynamic factor models in the time domain and its usefulness for dimension reduction in multiple time series. The core of this presentation is the work of Peña and Poncela (2006) and the recent extension for dynamic seasonal factors by Nieto, Peña and Saboya (2013). Then a new procedure for dimension reduction proposed by Peña and Yohai (2013) will be presented. It is shown that this procedure overcomes some of the problems of the dynamic principal components proposed by Brillinger (1981).