Analisis Runtun Waktu Untuk Memprediksi Tingkat Kemiskinan di Jawa Barat Dengan Model Double Exponential Smoothing (DES)

Abstract

This study aims to predict the poverty rate in West Java Province using time series analysis methods, specifically Holt's Double Exponential Smoothing (DES). Data were taken from the Central Statistics Agency (BPS) for the period 2015–2023. The results of the analysis show that the DES model is able to project the poverty rate with high accuracy, with a MAPE value of 9.97%. The projection shows an increase in the poverty rate from 8.38% in 2024 to 9.40% in 2028. This study emphasizes the importance of policy interventions based on the New Public Governance (NPG) theory to address this increasing trend. The recommended strategic steps include improving the quality of human resources through vocational education, strengthening MSMEs, and empowering the poor. The results of this study are expected to be a scientific reference for planning more effective social and economic policies in alleviating poverty in West Java.