Evaluation of container forecasting methods for analyzing port container terminal performance using agent-based simulation

Forecasting is a method to evaluate the performance prediction results of a future company, and in this paper 4 unvariate forecasting models are used for the number of container throughput volumes in a container terminal, where the four models are Moving Average (MA), Simple Exponential Smoothing (ES), Addictive and Multiplicative. And the method proposed for this paper is to prove the value of the prediction accuracy of the container throughput by applying data per day and will be comparing to obtain the measurement accuracy value of prediction result using Mean Square Error (MSE) and Root Mean Square Error (RMSE). It will get the smallest accuracy value as best prediction result from generating whole model for forecasting container throughput which will be implemented into Anylogic simulation as evaluation of event logs performance based on forecast result so that from result of anylogic forecast it will get performance time and cost in the form of event logs for the future time.