<![CDATA[This research aims to compare the performance of two prediction algorithms, XGBoost Regression and Random Forest Regression, in predicting poverty levels in the DKI Jakarta area. For this research, researchers obtained data from the DKI Jakarta Central Statistics Agency (BPS) covering the period 2010 to 2023. The testing method used involved measuring Mean Squared Error (MSE) and Mean Absolute Percentage Error (MAPE) to assess the accuracy of predictions from the two algorithms. . The findings show that the Random Forest Regression algorithm generally produces more accurate predictions compared to the XGBoost Regression algorithm as seen from the test results on (MSE) and (MAPE) for most of the areas analyzed. As with MAPE for the West Jakarta area, the test results for XGBoost Regression were 1.43, while Random Forest Regression produced 1.42, so Random Forest Regression is better than XGBosst Regression. However, in the Seribu Islands, the MAPE for XGBoost is better with a value of 4.49 than for Random Forest Regression which has a value of 4.56. Then MSE Random Forest is better than XGBoost in this prediction test. For example, in the Central Jakarta area with a value of 0.02 for XGBoost Regression, while Random Forest Regression has a smaller test result with a value of 0.01.]]>
Copyrights © 2024
