Parameter change test for random coefficient integer-valued autoregressive processes with application to polio data analysis. (English) Zbl 1221.62126

The first-order autoregressive model with integer-valued data is defined by the ‘thinning’ operator designated by F. W. Steutel and K. van Harn [Ann. Probab. 7, 893–899 (1979; Zbl 0418.60020)]. Let \(X\) be an integer-valued random variable and let \(\psi\in [0,1]\). The ‘thinning’ operator “\(\circ\)” is defined as \(\psi\circ X=\sum_{i=1}^{X}B_i\), where \(\{B_i\}\) is an independent of \(X\) i.i.d. Bernoulli random sequence with \(P(B_i=1)=\psi\). With this operator, the INAR(l) model [M.A. Al-Osh and A.A. Alzaid, J. Time Ser. Anal. 8, 261–275 (1987; Zbl 0617.62096)] is defined as \(X_t=\psi\circ X_{t-1}+Z_t,\;t\geq1,\) where \(\{Z_t\}\) is a sequence of independent of \(X_0\) i.i.d. non-negative integer-valued random variables with mean \(\lambda\) and variance \(\sigma^2_z\). The RCINAR(l) model introduced by H. Zheng, I.V. Basawa and S. Datta [J. Stat. Plann. Inference 137, No. 1, 212–229 (2007; Zbl 1098.62117)] is defined by the recursive equation \(X_t=\psi_t\circ X_{t-1}+Z_t,\;\geq1,\) where \(\{\psi_t\}\) is an i.i.d. sequence with cumulative distribution function defined on \([0,1)\), and \(\{Z_t\}\) is a sequence of non-degenerate and non-negative integer-valued random variables with \(E(Z_t^4)<\infty\). Let \(\psi=E(\psi_t)\), \(\sigma_{\psi}^2=\text{Var}(\psi_t)\), \(\lambda=E(Z_t)\), \(\tau^2=\psi^2=\sigma^2_{\psi}\), \(\sigma^2_z=\text{Var}(Z_t)\). The process \(\{X_t\}\) is an ergodic Markov chain and has a stationary distribution.
The authors deal with the problem of testing for a parameter change in a first-order random coefficient integer-valued autoregressive RCINAR(l) model. They employ the cumulative sum (CUSUM) test for detecting changes of the mean autoregressive rate \(\psi=E(\psi_t)\) and the mean innovation \(\lambda=E(Z_t)\) which is based on the conditional least-squares and modified quasi-likelihood estimators. It is shown that, under regularity conditions, the CUSUM test has the same limiting distribution as the supremum of the squares of independent Brownian bridges. The CUSUM test is then applied to the analysis of a monthly polio counts data set and demonstrates that there is a parameter change in the RCINAR(l) model.


62M10 Time series, auto-correlation, regression, etc. in statistics (GARCH)
62M07 Non-Markovian processes: hypothesis testing
62F03 Parametric hypothesis testing
62P10 Applications of statistics to biology and medical sciences; meta analysis


Full Text: DOI


[1] Alzaid, An integer-valued pth order autoregressive structure (INAR(p)) process, Journal of Applied Probability 27 pp 314– (1990) · Zbl 0704.62081
[2] Al-Osh, First order integer-valued autoregressive (INAR(1)) processes, Journal of Time Series Analysis 8 pp 261– (1987) · Zbl 0617.62096
[3] Cardinal, On the application of integer-valued time series models for the analysis of disease incidence, Statistics in Medicine 18 pp 2025– (1999)
[4] Chan, Parametric Statistical Change Point Analysis (2000) · doi:10.1007/978-1-4757-3131-6
[5] Chan, Monte Carlo EM estimation for time series models involving counts, Journal of the American Statistical Association 90 pp 242– (1995) · Zbl 0819.62069
[6] Csörg, Limit Theorems in Change-point Analysis (1997)
[7] Davis, On autocorrelation in a Poisson regression model, Biometrika 87 pp 491– (2000) · Zbl 0956.62075
[8] Du, The integer-valued autoregressive (INAR(p)) model, Journal of Time Series Analysis 12 pp 129– (1991) · Zbl 0727.62084
[9] Fahrmeir, Multivariate Statistical Modelling Based on Generalized Linear Models (1994) · Zbl 0809.62064 · doi:10.1007/978-1-4899-0010-4
[10] Franke, Developments in Time Series Analysis: In Honour of Maurice B. Priestley (1993)
[11] Ferguson, A Course in Large Sample Theory (1996) · doi:10.1007/978-1-4899-4549-5
[12] Jorgensen, A state space model for multivariate longitudinal count data, Biometrika 86 pp 169– (1999)
[13] Klimko, On conditional least squares estimation for stochastic processes, Annals of Statistics 6 pp 629– (1978) · Zbl 0383.62055
[14] Kuk, The Monte Carlo Newton Raphson algorithm, Journal of Statistical Computation and Simulation 59 pp 233– (1997) · Zbl 0912.62033
[15] Lee, Test for parameter change in stochastic processes based on conditional least-squares estimator, Journal of Multivariate Analysis 93 pp 375– (2005) · Zbl 1066.62082
[16] Lee, The cusum test for parameter change in time series models, Scandinavian Journal of Statistics 30 pp 781– (2003) · Zbl 1053.62085
[17] Page, A test for change in a parameter occurring at an unknown point, Biometrika 42 pp 523– (1955) · Zbl 0067.11602 · doi:10.1093/biomet/42.3-4.523
[18] Steutal, Discrete analogues of self-decomposability and stability, The Annals of Probability 7 pp 893– (1979)
[19] Zeger, A regression model for time series of counts, Biometrika 75 pp 621– (1988) · Zbl 0653.62064
[20] Zheng, Inference for pth-order random coefficient integer-valued autoregressive process, Journal of Time Series Analysis 27 pp 411– (2006)
[21] Zheng, The first order random coefficient integer-valued autoregressive processes, Journal of Statistical Planning and Inference 173 pp 212– (2007) · Zbl 1098.62117
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.