<![CDATA[This study analyzes and predicts the poverty rate in Manado City using Multiple Linear Regression (MLR) based on annual data from Statistics Indonesia (BPS). The dependent variable is the percentage of poor people, while the predictor variables include total population, average years of education, and Human Development Index (HDI). An alternative specification incorporates time trend controls. The analysis includes multicollinearity testing, OLS estimation, model diagnostics, and Leave-one-Out Cross Validation (LOOCV). Model A showed moderate explanatory power (R² = 0.4178) and good prediction accuracy (MAPE = 4.17%). Model B improved the behavior of the residuals by reducing autocorrelation and increasing the overall stability of the model. Education shows a negative and significant relationship with poverty, while the HDI coefficient requires careful interpretation due to multicollinearity. These findings suggest that expanding educational attainment and strengthening human development can effectively reduce poverty. Future research can integrate additional socioeconomic variables or adopt time series-based models to improve long-term predictive performance.]]>
Copyrights © 2025
