Geodesic
Some recent developments in Markov Chain Monte Carlo for cointegrated time series
Maciej Marowka1 ; Gareth W. Peters2 ; Nikolas Kantas1 ; Guillaume Bagnarosa3
1Imperial College London, UK
2University College London, UK
3ESC Rennes School of Business, France
ESAIM. Proceedings, Tome 59 (2017), pp. 76-103
Cet article a éte moissonné depuis la source EDP Sciences

Voir la notice de l'article

We consider multivariate time series that exhibit reduced rank cointegration, which means a lower dimensional linear projection of the process becomes stationary. We will review recent suitable Markov Chain Monte Carlo approaches for Bayesian inference such as the Gibbs sampler of [41] and the Geodesic Hamiltonian Monte Carlo method of [3]. Then we will propose extensions that can allow the ideas in both methods to be applied for cointegrated time series with non-Gaussian noise. We illustrate the efficiency and accuracy of these extensions using appropriate numerical experiments.
DOI : 10.1051/proc/201759076

Maciej Marowka  1   ; Gareth W. Peters  2   ; Nikolas Kantas  1   ; Guillaume Bagnarosa  3

1 Imperial College London, UK
2 University College London, UK
3 ESC Rennes School of Business, France 
@article{EP_2017_59_a5,
     author = {Maciej Marowka and Gareth W. Peters and Nikolas Kantas and Guillaume Bagnarosa},
     title = {Some recent developments in {Markov} {Chain} {Monte} {Carlo} for cointegrated time series},
     journal = {ESAIM. Proceedings},
     pages = {76--103},
     year = {2017},
     volume = {59},
     doi = {10.1051/proc/201759076},
     language = {en},
     url = {http://geodesic.mathdoc.fr/articles/10.1051/proc/201759076/}
}
TY  - JOUR
AU  - Maciej Marowka
AU  - Gareth W. Peters
AU  - Nikolas Kantas
AU  - Guillaume Bagnarosa
TI  - Some recent developments in Markov Chain Monte Carlo for cointegrated time series
JO  - ESAIM. Proceedings
PY  - 2017
SP  - 76
EP  - 103
VL  - 59
UR  - http://geodesic.mathdoc.fr/articles/10.1051/proc/201759076/
DO  - 10.1051/proc/201759076
LA  - en
ID  - EP_2017_59_a5
ER  - 
%0 Journal Article
%A Maciej Marowka
%A Gareth W. Peters
%A Nikolas Kantas
%A Guillaume Bagnarosa
%T Some recent developments in Markov Chain Monte Carlo for cointegrated time series
%J ESAIM. Proceedings
%D 2017
%P 76-103
%V 59
%U http://geodesic.mathdoc.fr/articles/10.1051/proc/201759076/
%R 10.1051/proc/201759076
%G en
%F EP_2017_59_a5
Maciej Marowka; Gareth W. Peters; Nikolas Kantas; Guillaume Bagnarosa. Some recent developments in Markov Chain Monte Carlo for cointegrated time series. ESAIM. Proceedings, Tome 59 (2017), pp. 76-103. doi: 10.1051/proc/201759076

Cité par Sources :