Time Series Modeling for Unemployment Prediction in Indonesia Using ARIMA (Autoregressive Integrated Moving Average) Approach

Abstract

This study aims to develop an effective time series model for predicting the unemployment rate in Indonesia using the ARIMA (Autoregressive Integrated Moving Average) approach. The analysis was conducted using historical unemployment data from February 2010 to February 2020. After evaluating various ARIMA model configurations, the study identified the ARIMA (2,0,2) as the best-performing model. This model demonstrated superior forecasting accuracy, with an RMSE of 0.1629445, MAE of 0.1376129, and MAPE of 2.317308, along with a high R-squared value. The results of the ARIMA (2,0,2) model forecast for February 2021 indicate an unemployment rate of 5.070741, with a lower bound of 4.59359 and an upper bound of 5.547891. These findings provide valuable insights for policymakers and labor market stakeholders in Indonesia, enabling them to make informed decisions and implement targeted strategies to address unemployment challenges more effectively. The successful application of the ARIMA approach in this study highlights its potential as a robust and reliable tool for unemployment forecasting in the Indonesian context, contributing to the existing literature on time series modeling for labor market analysis.