We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
TÜRKİYE LİMANLARINDAKİ KONTEYNER VE YÜK ELLEÇLEME HACİMLERİNİN NARX SİNİR AĞLARI MODELİ İLE ÖNGÖRÜLMESİ.
- Authors
KILINÇ, Gamze; KARAATLI, Meltem; ÖMÜRBEK, Nuri
- Abstract
Purpose: In this study, in order to use the existing equipment more effectively and efficiently, the amount of container and cargo handling realized at the ports in Turkey has been predicted. The forecast of container and cargo volumes will enable the planning of container and cargo flow and thus support the improvement of the service quality of the enterprises. Methodology: In this study, the amount of container and cargo handling in ports in Turkey is forecasted with Artificial Neural Networks (ANN) Nonlinear External Input Autoregressive Network Model (NARX). In order to forecast the amount of container handling and cargo handling, which are two dependent variables in this study; The Foreign Exchange Basket Rate (USD-EURO), Gross Domestic Product (GDP), Consumer Confidence Index, Brent Oil, Export, Import and Industrial Production Index independent variables are used. Data set includes the monthly data for the period January 2004-October 2020. Findings: For the period July 2020-December 2021, 18-month forecast values of container and cargo handling quantities are calculated. Forecast and actual value comparisons are made for the first 4 months with announced and predictive values. In addition, MAPE, MAE, RMSE and MAD performance measures are calculated. Originality: Different from previous forecasting techniques, NARX neural network model is used for the prediction of container and cargo handling quantities in Turkish ports.
- Publication
Verimlilik Dergisi, 2022, Issue 2, p251
- ISSN
1013-1388
- Publication type
Article
- DOI
10.51551/verimlilik.896525