本研究提出一套結合季節性自我迴歸整合移動平均模式(Seasonal Autoregressive Integrated Moving Average, SARIMA)模式與類神經網路的組合預測模式用以作為機場短期貨物運量預測使用。本研究以桃園機場貨運運量為研究對象，研究期間為1993年1月至2013年4月。本研究使用平均絕對誤差(Mean Absolute Error, MAE)與平均絕對誤差百分比(Mean Absolute Percentage Error, MAPE)兩種指標來衡量SARIMA模式、類神經網路與組合預測模式的預測績效。研究結果發現，三種預測模式之MAPE皆小於10%，而本研究所提出之組合預測模式，較單獨使用單一模式更為精確，且能提高 預測能力，將能有效提供政府當局作為機場短期貨物運量預測的參考使用。

In this research, a novel approach combining the Seasonal Autoregressive Integrated Moving Average (SARIMA) and the Artificial Neural Network (ANNs) model are proposed to forecast the air cargo volume in Taiwan Taoyuan International airport. Monthly time series data covering from January 1993 toApril 2013 is used in this research. The Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) are used to compare the performance of the combined model against other two models (i.e., the SARIMA model and the ANNs model). Results show that the MAPE of three models are all below 10% and the proposed combined forecast model is more accurate than single model. The proposed combined forecast model can be used for effective short-term air cargo volume forecast for the government.