FIT OF STATISTICAL FORECASTING MODEL BERDASARKAN VARIABEL ANGKA KEMISKINAN DI PROVINSI KEPULAUAN BANGKA BELITUNG

Abstract

Poverty is one of the main problems in economic development and is considered to be a variable to measure the success of the economic development of a region. This study is limited to the analysis and determination of the best forecasting statistical model for the variable poverty rate in the Bangka Belitung Islands Province area based on R Square, Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE) assessments. This study uses the Exponential Smoothing forecasting method which emphasizes the procedure of continuous improvement of the latest observation objects which hopefully can provide the appropriate results. In general, the double exponential smoothing model from Holt's is the best projection model compared to other exponential smoothing models for projecting poverty data in the Bangka Belitung Islands Province with historical data for 2002-2018 with an increase in projections in 2019 of 0.37 % with Upper Criteria Limit (UCL) of 1.07% and Lower Criteria Limit (LCL) of -0.33% with a value of R Square of 0.627 which means that the independent variable can explain the variance of the dependent variable of 62.7% of this model, and the value of RMSE is 0.328 and MAPE is 22.162. The results of this model when compared to other models have relatively larger R Squared values ??and smaller RMSE and MAPE values.