Forecasting the number of mobile phone users in Thailand

Abstract

The purpose of this study was to create a predictive model. Compare models for forecasting the number of mobile phone users in Thailand. and to forecast the number of mobile phone users in the 1st quarter of 2023–4th quarter of 2023. The technique used to forecast the number of mobile phone users in Thailand is the exponential smoothing method. Holt's exponential smoothing methods, simple regression analysis, and Brown's exponential smoothing method, which the data studied are serial data. quarterly time from the 1st quarter of 2002 to the 4th quarter of 2022 from the data.go.th website, 84 data points, divided into two sets. Set 1 from the 1st quarter of 2002—2002 quarter. 4th year 2021, 80 data sets, 2nd set from 1st quarter 2022 to 4th quarter 2022, 4 data sets for comparing the accuracy of forecasting methods using the percentage of accuracy criteria. Expected average absolute movement The analysis results found that Brown's exponential smoothing method has a MAPE=0.26, which is less than 10%, therefore falling within the criteria for high accuracy in forecasting. Brown's exponential smoothing method is a suitable model for forecasting, which has a lower mean absolute percentage error (MAPE) than Holt's exponential smoothing method and the simple regression analysis method. The model is where m represents the number of periods to be forecasted forward. The default value was the 1st quarter of 2023; m=1 in forecasting the number of mobile phone subscribers in Thailand. The forecast values will be as follows: 120,472,244, 120,403,775, 120,335,307, and 120,266,839 people. We were forecasting the number of mobile phone users in Thailand in advance. It shows that the number of mobile phone users in Thailand tends to decrease.