Poverty Line Forecasting Model Using Double Exponential Smoothing Holt-s Method
Roni Aminudin;Yeffry Handoko Putra
Magister Sistem Informasi, Universitas Komputer Indonesia, Jl. Dipatiukur No. 112-116, Bandung, Jawa Barat
Abstract
Abstract. This research aims to forecast the Poverty Line, to help a government obtain accurate and fast information. The method used in this research is Double Exponential Smoothing Holt-s Method. This method is a part of the data based time series analysis. The research applies the forecasting theory to produce a poverty line forecast for the coming year. Next, this research is analyzing data patterns and determine the best parameter values. Double Exponential Smoothing Holt-s method uses the parameters Alpha (α) and Gamma (γ). To determine the best parameter value is to use the trial and error method. The best parameter value produces the smallest value of MAPE (Mean Absolute Percentage Error). The data pattern shows the trend, meaning that the Double Exponential Smoothing Holt-s method is appropriate for use in this research. The parameter values generated from the trial and error methods are Alpha (α) of 0.7 and Gamma (γ) of 0.1, which produced the smallest measure of accuracy, in this research using MAPE. By observing the results of the forecasting that has been done, this forecasting model has a very good performance. Poverty Line value will keep increasing, in accordance with increasing consumption patterns and rising prices of basic necessities.
Keywords: Poverty Line, Forecasting, Double Exponential Smoothing Holt-s Method, Time Series, MAPE (Mean Absolute Percentage Error)
Topic: Informatic and Information System