Abstract
Modeling of high dimensional time series by linear time series models such as vector autoregressive models is often marred by the so-called “curse of dimensionality”. In order to overcome this problem generalized linear dynamic factor models (GDFM's) maybe used. In high-dimensional time series the single univariate time series are often sampled at different frequencies. This is the so-called mixed-frequency situation. We consider identifiability of the underlying high-frequency GDFM (i.e. the GDFM generating the data at the highest sampling frequency occurring) in the case of mixed frequency data and we shortly describe two estimation procedures in this situation based on the EM algorithm.
Original language | English |
---|---|
Pages (from-to) | 1008-1013 |
Number of pages | 6 |
Journal | 18th IFAC Symposium on System Identification SYSID 2018: Stockholm, Sweden, 9-11 July 2018 |
Volume | 51 |
Issue number | 15 |
DOIs | |
Publication status | Published - 1 Jan 2018 |
Externally published | Yes |