A note on objective Bayes analysis for graphical vector autoregressive models

Risultato della ricerca: Contributo in libroContributo a convegno

Abstract

Vector Autoregressive (VAR) models are widely used to estimate and forecast multivariate time series. However, the large number of parameters of VAR models can lead to unstable inference and inaccurate forecasts, particularly with many variables. For this reason, restrictions supported by the data are usually required.We propose an objective Bayes approach based on graphical VAR models for learning contemporaneous dependencies as well as dynamic interactions among variables. We show that, if the covariance matrix at each time is Markov with respect to the same decomposable graph, then the likelihood of a graphical VAR can be factorized as an ordinary decomposable graphical model. Additionally, using a fractional Bayes factor approach, we are able to obtain the marginal likelihood in closed form and perform Bayes graphical model selection with limited computational burden.
Lingua originaleEnglish
Titolo della pubblicazione ospiteBook of short Papers SIS 2018
Pagine1580-1585
Numero di pagine6
Stato di pubblicazionePubblicato - 2018
EventoSIS2018: 49th Scientific Meeting of the Italian Statistical Society - Palermo
Durata: 20 giu 201822 giu 2018

Convegno

ConvegnoSIS2018: 49th Scientific Meeting of the Italian Statistical Society
CittàPalermo
Periodo20/6/1822/6/18

Keywords

  • Bayesian model selection
  • decomposable graphical model
  • fractional Bayes factor
  • multivariate time series

Fingerprint

Entra nei temi di ricerca di 'A note on objective Bayes analysis for graphical vector autoregressive models'. Insieme formano una fingerprint unica.

Cita questo