Forecasting Ticket Sales – the Case of Commuter Rail in South Africa

Abstract

The application of quantitative statistical modelling in the commuter rail environment is explored in this research paper. The research explored the application of various time series models as well as the ARIMA model and regression analysis. The application of two forecast combinations was also explored to improve the accuracy of the forecasts. The ARIMA model in combination with the seasonal decomposition was used to forecast the data for a period of 18 months.