Multivariate Count Data Models for Time Series Forecasting

Autor:	Michael Eichler, Yuliya Shapovalova, Nalan Basturk
Přispěvatelé:	QE Econometrics, RS: GSBE Theme Data-Driven Decision-Making, RS: FSE DACS Mathematics Centre Maastricht, RS: GSBE other - not theme-related research, Weiss, C.H.
Rok vydání:	2021
Předmět:	Heteroscedasticity Multivariate statistics state-space model Computer science Science QC1-999 General Physics and Astronomy Inference multivariate count data Astrophysics 01 natural sciences Article Count data models 010104 statistics & probability transactions REGRESSION 0502 economics and business Statistics 0101 mathematics Time series bank failures 050208 finance State-space representation Physics Data Science 05 social sciences QB460-466 Autoregressive model INGACRCH POISSON Count data
Zdroj:	Entropy, 23(6):718. Multidisciplinary Digital Publishing Institute (MDPI) Entropy Volume 23 Issue 6 Weiss, C.H. (ed.), Time Series Modelling, pp. 317-340 Entropy, Vol 23, Iss 718, p 718 (2021) Entropy, 23, 1-23 Entropy, 23, 6, pp. 1-23
ISSN:	1099-4300
Popis:	Count data appears in many research fields and exhibits certain features that make modeling difficult. Most popular approaches to modeling count data can be classified into observation and parameter-driven models. In this paper, we review two models from these classes: the log-linear multivariate conditional intensity model (also referred to as an integer-valued generalized autoregressive conditional heteroskedastic model) and the non-linear state-space model for count data. We compare these models in terms of forecasting performance on simulated data and two real datasets. In simulations, we consider the case of model misspecification. We find that both models have advantages in different situations, and we discuss the pros and cons of inference for both models in detail.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_dedup___::e0107e368f21b59c1881e123c901124d https://doi.org/10.3390/e23060718 Zobrazit plný text záznamu Plný text ve formátu PDF Plný text ve formátu HTML
Nepřihlášeným uživatelům se plný text nezobrazuje	K zobrazení výsledku je třeba se přihlásit.