Heteroscedasticity modeling in transportation engineering is primarily conducted in short-term traffic condition forecasting to generate time varying prediction intervals around the point forecasts through quantitatively predicting the conditional variance of traffic condition series. Until recently, the generalized autoregressive conditional heteroscedasticity (GARCH) model and the stochastic volatility model have been two major approaches adopted from the field of financial time series analysis for traffic heteroscedasticity modeling. In this paper, recognizing the pronounced seasonal pattern in traffic condition data, a simple seasonal adjustment approach is explored for modeling seasonal heteroscedasticity in traffic-flow series, and four types of seasonal adjustment factors are proposed with respect to daily or weekly patterns. Using real-world traffic-flow data collected from highway systems in the United Kingdom and the United States, the proposed seasonal adjustment approach is implemented and validated. Empirical results show that the proposed model can effectively capture and hence model the seasonal heteroscedasticity in traffic-flow series. In addition, through a comparison with the conventional GARCH model, the proposed approach is shown to consistently generate improved performances in terms of prediction interval construction. Potential applications are discussed to explore the value of heteroscedasticity modeling in transportation engineering studies.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Modeling Seasonal Heteroscedasticity in Vehicular Traffic Condition Series Using a Seasonal Adjustment Approach


    Beteiligte:
    Shi, Guogang (Autor:in) / Guo, Jianhua (Autor:in) / Huang, Wei (Autor:in) / Williams, Billy M. (Autor:in)


    Erscheinungsdatum :

    24.02.2014




    Medientyp :

    Aufsatz (Zeitschrift)


    Format :

    Elektronische Ressource


    Sprache :

    Unbekannt