Adrian Raftery: Time Series and Point Process Research

Much of my time series work has focused on non-Gaussian time series. I have found a fruitful approach to this to be the Mixture Transition Distribution (MTD) model introduced by Raftery (1985); for a review of this class of models see Berchtold and Raftery (2002). This replaces the standard predictive distribution which is often taken to be Gaussian centered around a linear combination of past values, with a mixture of distributions that depend on each of the past values. This makes it easy to model non-Gaussian time series, such as discrete-valued, positive-valued, compositional, or time series of angles. It also provides an approach to robust time series analysis, by modeling outliers explicitly (Le et al 1996b). It can be estimated using the freely available GMTD software, written by Andre Berchtold.

I have also worked on estimation methods for long-memory time series models (Haslett and Raftery 1989). Software developed by Chris Fraley and myself for fast approximate maximum likelihood estimation of long memory time series models is available in Splus under the name arima.fracdiff and also at fracdiff for S, and for Fortran at fracdiff for Fortran. Most recently I have been working on online prediction under model uncertainty. Raftery et al (2007) develops an extension of Bayesian model averaging to the time series case, called Dynamic Model Averaging (DMA). This is related to my work on probabilistic weather forecasting using model ensembles.

My interest in (temporal) point processes has focused on detecting and estimating change points (for a review see Raftery (1994)), and on applications to software reliability (Raftery 1987, 1988).

For my work on spatial point processes, click here.

Papers

Young, W.C., Yeung, K.Y. and Raftery, A.E. (2019). Identifying dynamical time series model parameters from equilibrium samples, with application to gene regulatory networks. Statistical Modelling 19:444--465.

Onorante, L. and Raftery, A.E. (2016). Dynamic Model Averaging in Large Model Spaces. European Economic Review 81:2-14.

McCormick, T.M., Raftery, A.E., Madigan, D. and Burd, R.S. (2012). Dynamic Logistic Regression and Dynamic Model Averaging for Binary Classification. Biometrics 68:23-30.

Raftery, A.E., Karny, M., and Ettler, P. (2010). Online Prediction Under Model Uncertainty Via Dynamic Model Averaging: Application to a Cold Rolling Mill. Technometrics 52:52-66.

Berchtold, A. and Raftery, A.E. (2002). The Mixture Transition Distribution (MTD) model for high-order Markov chains and non-Gaussian time series. Statistical Science, 17, 328-356.

Le, N.D., Raftery, A.E. and Martin, R.D. (1996a). Robust order selection in autoregressive models using robust Bayes factors. Journal of the American Statistical Association, 91, 123-131.

Le, N.D., Martin, R.D. and Raftery, A.E. (1996b). Modeling outliers, bursts and flat stretches in time series using mixture transition distribution (MTD) models. Journal of the American Statistical Association, 91, 1504-1515.

Raftery, A.E. (1994). Change point and change curve modeling in stochastic processes and spatial statistics. Journal of Applied Statistical Science, 1, 403-424. Earlier technical report version.

Raftery, A.E. and Tavare, S. (1994). Estimation and modelling repeated patterns in high-order Markov chains with the mixture transition distribution (MTD) model. Journal of the Royal Statistical Society, series C - Applied Statistics, 43, 179-200.

Grunwald, G.K., Raftery, A.E. and Guttorp, P. (1993a). Time series of continuous proportions. Journal of the Royal Statistical Society, series B, 55, 103-116.

Grunwald, G.K., Guttorp, P. and Raftery, A.E. (1993b). Prediction rules for exponential family state-space models. Journal of the Royal Statistical Society, series B, 55, 937-943.

Stephen, E., Raftery, A.E. and Dowding, P. (1990). Forecasting spore concentrations: A time series approach. International Journal of Biometeorology, 34, 87-89.

Raftery, A.E. (1989). Are ozone exceedance rates decreasing? Statistical Science, 4, 378-381.

Haslett, J. and Raftery, A.E. (1989). Space-time modelling with long-memory dependence: Assessing Ireland's wind power resource (with Discussion). Journal of the Royal Statistical Society, series C - Applied Statistics, 38, 1-50.

Raftery, A.E. (1988). Analysis of a simple debugging model. Journal of the Royal Statistical Society, series C - Applied Statistics, 37, 12-22.

Raftery, A.E. (1987). Inference and prediction for a general order statistic model with unknown population size. Journal of the American Statistical Association, 82, 1163-1168.

Martin, R.D. and Raftery, A.E. (1987). Robustness, computation, and non-Euclidean models. Journal of the American Statistical Association, 82, 1044-1050.

Akman, V.E. and Raftery, A.E. (1986). Bayes factors for non-homogeneous Poisson processes with vague prior information. Journal of the Royal Statistical Society, series B, 48, 322-329.

Raftery, A.E. (1986). A note on Bayes factors for log-linear contingency table models with vague prior information. Journal of the Royal Statistical Society, series B, 48, 249-250.

Akman, V.E. and Raftery, A.E. (1986). Asymptotic inference for a change-point Poisson process. Annals of Statistics, 14, 1583-1590.

Raftery, A.E. and Akman, V.E. (1986). Bayesian analysis of a Poisson process with a change-point. Biometrika, 73, 85-89.

Raftery, A.E. (1985). A model for high-order Markov chains. Journal of the Royal Statistical Society, series B, 47, 528-539.

Raftery, A.E. (1985). Invited review: Time series analysis. European Journal of Operational Research, 20, 127-137.

Raftery, A.E., Haslett, J. and McColl, E. (1982). Wind power: a space-time process? In Time series analysis: theory and practice 2 (O.D. Anderson, ed.), North-Holland, pp. 191-202.

Raftery, A.E. (1982). Generalised non-normal time series models. In Time series analysis: theory and practice 1 (O.D. Anderson, ed.), North-Holland, pp. 621-640.

Raftery, A.E. (1980). Estimation efficace pour un processus autoregressif exponentiel a densite discontinue. Publications de l'Institut de Statistique des Universites de Paris, 25, 64-91.

These papers are being made available here to facilitate the timely dissemination of scholarly work; copyright and all related rights are retained by the copyright holders.

Updated April 13, 2020.