Periodic autoregressive model identification using genetic algorithms

Abstract

Periodic autoregressive (PAR) models extend the classical autoregressive models by allowing the parameters to vary with seasons. Selecting PAR time-series models can be computationally expensive, and the results are not always satisfactory. In this article, we propose a new automatic procedure to the model selection problem by using the genetic algorithm. The Bayesian information criterion is used as a tool to identify the order of the PAR model. The success of the proposed procedure is illustrated in a small simulation study, and an application with monthly data is presented.

Seasonal and Periodic Autoregressive Time Series Models Used for Forecasting Analysis of Rainfall Data

“…Tesfaye, Meerschaert and Anderson [33] has applied this periodic ARMA modelling on monthly river flows. Recently Ursu E. et.al have used PAR model identification using genetic algorithms and also have done work on application of periodic autoregression process to the modelling of the Garonne river flows [36], [37]. We also find the various application of PAR model on river flow [33] and stream flow [29], [40].…”

Modelling and forecasting the trends of life cycle curves in the production of non-renewable resources

“…[34]. In recent years, it has received significant attention, particularly with reference to the problem of curve-fitting in time series analysis [35].…”

Conditional Granger Causality and Genetic Algorithms in VAR Model Selection

“…Subsequently, the multivariate case was tackled, with applications in reducing model complexity for those cases where the total number of calculations required for estimation renders the research effort computationally expensive. Specifically, VAR model identification was proved to be efficient through testing in simulation studies, where a synthetic dataset was generated using a predetermined model [34]. Useful economical applications for the evolutionary VAR methodology were mainly orientated towards the manipulation of complex environments for stock market trading systems or commodities markets forecasting [35,36].…”