Konteyner taşımacılığının, denizyolu ticaretindeki önemi her geçen gün artmaktadır. Konteyner hacminin etkili tahmini ise liman planlaması ve işletimi için bir karar desteği sağlamaktadır. Bu nedenle liman yönetimlerinin geleceğe yönelik planları açısından tahminleme çalışmaları önemli bir rol oynamaktadır. Bu çalışmada, Antalya’da bulunan Port Akdeniz Limanı için yapılan tahmin modellerinde Ocak 2008-Aralık 2017 (120 ay) dönemi konteyner istatistikleri veri seti olarak kullanılmıştır. Liman işletmesinin yük talep tahmini, konteyner bazında ve mevsimsel farklılıklar dikkate alınarak, Ocak 2018-Aralık 2019 (24 ay) dönemi için yapılmıştır. Gri Tahmin ve Box-Jenkins yöntemlerinin kullanıldığı çalışmada, konteyner tahminleri Gri Model (1,1) ve ARIMA (0,1,0)x(0,1,1)12 modelleri ile analiz edilmiştir. Tahmin sonuçları başarı kriterleri ile değerlendirildiğinde, Gri Model (1,1)’in MAPE ve MAE değerlerinin daha düşük olduğu gözlemlenmiştir. Ancak hem RMSE ve MSE hem de sapma değerleri dikkate alındığında ise; ARIMA (0,1,0)x(0,1,1)12 modelinin daha iyi ve uygun tahmin değerleri verdiği tespit edilmiştir.
The importance of container transportation in the maritime trade is increasing every day. The effective estimate of the container volume provides a decision support for port planning and operation. Therefore, the work of prediction in terms of the future plans of the port management plays an important role. In this study, the forecast models for Port Mediterranean Port located in Antalya used container statistics for the period January 2008–December 2017 (120 months) as a data set. The cargo demand for the port operations is forecasted, based on containers and taking into account seasonal differences, for the January 2018-December 2019 (24 months) period. In the study using the Grey Prediction and Box-Jenkins methods, container predictions were analyzed using the Grey Model (1,1) and the ARIMA (0,1,0)x(0,1,1)12 models. When the forecast results were evaluated by success criteria, it was observed that the MAPE and MAE values of the Grey Model (1,1) were lower. However, taking into account both RMSE and MSE and deviation values, the ARIMA (0,1,0)x(0,1,1)12 model has been found to provide better and more appropriate predictive values.
The importance of container transportation in maritime trade is increasing day by day. The effective prediction of container volume provides decision support for planning and operations of the ports. Therefore, forecasting are crucially important for the future plans of port management have an important place. In this study, container statistics for January 2008-December 2017 (120 months) period were used in the estimation models for Port Akdeniz Port in Antalya. The freight demand forecast of the port management is made for the period of January 2018-December 2019 (24 months), taking into consideration the seasonal differences on the basis of containers. In the study using Gray Estimation and Box-Jenkins (B-J) methods, container volumes estimations were carried out using Gray Model (1,1) and ARIMA (0,1,0)x(0,1,1)12 models. As estimation results are evaluated with success criteria, it is observed that MAPE and MAE values of Gray Model (1,1) are lower. However, considering both RMSE and MSE and deviation values; it is determined that ARIMA (0,1,0)x(0,1,1)12 model gives better and more suitable estimation values.
Alan : Sosyal, Beşeri ve İdari Bilimler
Dergi Türü : Uluslararası
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|