On a mixture autoregressive model

We generalize the Gaussian mixture transition distribution (GMTD) model introduced by Le and co‐workers to the mixture autoregressive (MAR) model for the modelling of non‐linear time series. The models consist of a mixture of K stationary or non‐stationary AR components. The advantages of the MAR model over the GMTD model include a more full range of shape changing predictive distributions and the ability to handle cycles and conditional heteroscedasticity in the time series. The stationarity conditions and autocorrelation function are derived. The estimation is easily done via a simple EM algorithm and the model selection problem is addressed. The shape changing feature of the conditional distributions makes these models capable of modelling time series with multimodal conditional distributions and with heteroscedasticity. The models are applied to two real data sets and compared with other competing models. The MAR models appear to capture features of the data better than other competing models do.

[1]  C. Elton,et al.  The Ten-Year Cycle in Numbers of the Lynx in Canada , 1942 .

[2]  P. Moran.  The Statistical Analsis of the Canadian Lynx cycle. 1. Structure and Prediction. , 1953 .

[3]  V. E. Beneš.  Existence of finite invariant measures for Markov processes , 1967 .

[4]  Gwilym M. Jenkins,et al.  Time series analysis, forecasting and control , 1972 .

[5]  H. Akaike,et al.  Information Theory and an Extension of the Maximum Likelihood Principle , 1973 .

[6]  Dean W. Wichern,et al.  Changes of Variance in First-Order Autoregressive Time Series Models -With an Application , 1976 .

[7]  R. Davies.  Hypothesis testing when a nuisance parameter is present only under the alternative , 1977 .

[8]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[9]  G. Schwarz.  Estimating the Dimension of a Model , 1978 .

[10]  D. P. Gaver,et al.  First-order autoregressive gamma sequences and point processes , 1980, Advances in Applied Probability.

[11]  T. Louis.  Finding the Observed Information Matrix When Using the EM Algorithm , 1982 .

[12]  R. Engle.  Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation , 1982 .

[13]  V. Yohai,et al.  4 Robustness in time series and estimating ARMA models , 1985 .

[14]  A. J. Lawrance,et al.  Modelling and Residual Analysis of Nonlinear Autoregressive Time Series in Exponential Variables , 1985 .

[15]  V. Yohai,et al.  Influence Functionals for Time Series , 1986 .

[16]  T. Bollerslev,et al.  Generalized autoregressive conditional heteroskedasticity , 1986 .

[17]  C. Granger,et al.  Co-integration and error correction: representation, estimation and testing , 1987 .

[18]  Xiao-Li Meng,et al.  Using EM to Obtain Asymptotic Variance-Covariance Matrices: The SEM Algorithm , 1991 .

[19]  Vance L. Martin.  THRESHOLD TIME SERIES MODELS AS MULTIMODAL DISTRIBUTION JUMP PROCESSES , 1992 .

[20]  C. Granger,et al.  Modelling Nonlinear Economic Relationships , 1995 .

[21]  Z. D. Feng,et al.  Using Bootstrap Likelihood Ratio in Finite Mixture Models , 1994 .

[22]  J. Pemberton,et al.  Mixture models for time series , 1995, Journal of Applied Probability.

[23]  A. Raftery,et al.  Modeling flat stretches, bursts, and outliers in time series using mixture transition distribution models , 1996 .

[24]  On testing the number of components in finite mixture models with known relevant component distributions , 1997 .

[25]  P. Green,et al.  Corrigendum: On Bayesian analysis of mixtures with an unknown number of components , 1997 .

[26]  P. Green,et al.  On Bayesian Analysis of Mixtures with an Unknown Number of Components (with discussion) , 1997 .

[27]  E. Gassiat,et al.  The estimation of the order of a mixture model , 1997 .

[28]  Statistical inference for some nonlinear time series models , 1998 .