MODELLING OF ORDINAL TIME SERIES BY PROPORTIONAL ODDS MODEL
Abstract
Categorical time series data with random time dependent covariates often arise when the variable categories are assigned as categorical. There are several other models that have been proposed in the literature for the analysis of categorical time series. For example, Markov chain models, integer autoregressive processes, discrete ARMA models can be utilized for modeling of categorical time series. In general, the choice of model depends on the measurement of study variables: nominal, ordinal and interval. However, regression theory is successful approach for categorical time series which is based on generalized linear models and partial likelihood inference. One of the models for ordinal time series in regression theory is proportional odds model. In this study, proportional odds model approach to ordinal categorical time series is investigated based on a real air pollution data set and the results are discussed.
Keywords
Categorical time series, Ordinal regression, Proportional odds model
References
- Agresti, A. (2002). Categorical Data Analysis, 2nd ed., New Jersey:John Wiley.
- Bru, N., Despres, L. and Paroissin, C.A. Com- parison of Statistical Models for Short Categorical or Ordinal Time Series with Applications gy,arxiv.math/0702706v1. in Ecolo
- Fokianos, K. and Kedem, B. (2003). Regression Theory for Categorical Time Series, Sta- tistical Science. 18, 3, 357-376.
- Jacobs, P. and Lewis, P. (1978). Discrete Time Series Generated by Mixtures i: Correla- tion and Runs Properties, Journal of The Royal Statistical Society (Series B), 40(1), 94-105.
- Liu, L. and Agresti, A. (2005). The Analysis of Ordered Categorical data: An Overview and A Survey of Recent Developments, Sociedad de Estadıstica e Investigacion Operativa Test, 14, 1-73.
- McCullagh, P. (1980). Regression Models for Ordinal Data, Journal of the Royal Statis- tical Society - Series B 42, 109-142.
- McCullagh, P. (2005). The proportional Odds model, The Encyclopedia of Biostatistics (Editor: Armitage, P., Colton, T.), Wiley, NewYork.
- SPSS Inc. Released 2008. SPSS Statistics for Windows, Version 17.0. Chicago: SPSS Inc.
- www.havaizleme.gov.tr 09/08/2011) (Erişim Tarihi