A comparison of forecasting the results of road transportation needs

Abstract

Determining the size and quality of transport needs would not be possible without adequate forecasting based on the sales volume or demand for this service from the past periods. Traditional forecasting methods use econometric models that may be subject to serious errors. The use of the methods taking into account the variability of the studied phenomena or more advanced mathematical methods enables to minimize the error. Various methods of artificial intelligence such as a neural network, fuzzy sets, genetic algorithms, etc., have been recently successfully applied. The aim of this paper is to compare three forecasting methods that can be used for predicting the volume of road freight. The article deals with the effectiveness of three prediction methods, namely Winter's method for seasonal problems – a multiplicative version, harmonic analysis and harmonic analysis aided by the artificial immune system. The effectiveness of prediction was counted using MAPE errors (main average percentage error). The results of calculations were compared and the best example was presented.

First Published Online: 30 Mar 2012