Forecasting domestic tourism : application to Johor tourism data

Abstract

Serious tourism forecasting spans over three decades. A number of tourism forecasters have noted that complex forecasting models are seldom more accurate in predicting the future than simple ones. Hence, in this study, it shows through a Johor tourism forecasting comparison using the naïve model, moving average and exponential smoothing. An approach to selecting the method appropriate to Johor tourism data is based on criteria which are Mean Absolute Percentage Error (MAPE) and Mean Square Error (MSE). The naïve model, moving average and exponential smoothing are relatively simple to apply, requiring no more than a data series and a computer spreadsheet program. This study also explores the need of Tabu Search method in implying forecast. Tabu Search will be used to ensure the pattern of the studied data by searching the value of weighted constants. The confirmation is performed by Automated Computerized Forecasting (ACF) System.