Comparative Study of the ARIMA Method and Multiple Linear Regression in Metro City Population Growth Projections

Abstract

This study aims to compare the effectiveness of the ARIMA (Autoregressive Integrated Moving Average) method and multiple linear regression in projecting population growth in Metro City, Lampung. The analysis utilizes population data from 2010 to 2022, sourced from the Central Statistics Agency and the Population and Civil Registration Office. The methodologies employed include ARIMA modelling and multiple linear regression, with model evaluation conducted using metrics such as Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE). The findings indicate that the multiple linear regression model predicts an average population growth of 2,200 individuals per year, resulting in a total projection of 185,032 by 2030. In contrast, the ARIMA (2,1,1) model forecasts a total population of 169,500 for the same year. The conclusion drawn from this research suggests that while both methods possess distinct advantages, ARIMA is more effective in capturing seasonal patterns and long-term trends, whereas multiple linear regression offers greater interpretability. This study recommends the complementary use of both methods to enhance the accuracy of population growth projections.

Keywords

• population growth
• multiple linear regression
• arima
• population projection
• development planning