PoARX models for count time series

Jamie Halliday, Georgi N. Boshnakov

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

    160 Downloads (Pure)

    Abstract

    This paper introduces multivariate Poisson autoregressive models with exogenous covariates (PoARX) for modelling multivariate time series of counts. We state conditions for a PoARX process to be stationary and ergodic before proposing a computationally efficient procedure for estimation of parameters by the method of inference functions (IFM) and stating asymptotic normality of these estimators. Lastly, we demonstrate an application to count data for the number of people entering and exiting a building, and show how the different aspects of the model combine to produce a strong predictive model. We conclude by listing directions for future work.
    Original languageEnglish
    Title of host publicationITISE 2018 International Conference on Time Series and Forecasting. Proceedings, Granada, 19-21 September, 2018
    EditorsOlga Valenzuela, Fernando Rojas, Héctor Pomares, Ignacio Rojas
    Place of PublicationGranada (Spain)
    Pages1519-1530
    Number of pages12
    Publication statusPublished - 21 Sep 2018

    Keywords

    • Multivariate time series
    • Count data
    • Prediction
    • Copula

    Fingerprint

    Dive into the research topics of 'PoARX models for count time series'. Together they form a unique fingerprint.

    Cite this