우리나라 택배 물동량 예측 모형에 대한 연구

초록

[Purpose] Despite the fact that courier service volume is a key indicator for calculating the commission fee for courier companies, there is no scientific prediction study, so used simply an average growth rate index. The purpose of this study is to establish and verify models to estimate courier service volume affected by seasonal factors by using time series model. [Methodology] We used the single exponential smoothing method, Winter’s additive model, and seasonal ARIMA model, which are highly utilized among the various models used in the time series analysis, to estimate and compare the actual courier service volume with estimated volume. [Findings] This study found that the seasonal ARIMA model was the best, in particular, short－term forecasts within one year showed better prediction accuracy than other models. In comparative long－term forecasts over one year and less than two year, the Winter’s additive model, which considers trends and seasonality, can be effectively used. [Implications] This study statistically verified that even if the Winter’s additive model is less sophisticated compared with he seasonal ARIMA model, it can be used without no problems in decision making in reality.

키워드