<![CDATA[Rainfall is a crucial meteorological parameter that significantly affects various sectors, particularly in tropical regions such as Jambi City. However, daily rainfall data often contain outliers and imbalanced class distributions, which can degrade the performance of machine learning-based classification models. This study aims to conduct a preliminary analysis of the performance of several machine learning algorithms for daily rainfall classification in Jambi City by examining the effects of outlier removal. The algorithms evaluated include Support Vector Machine (RBF), K-Nearest Neighbor, Naive Bayes, Decision Tree, and Random Forest. Model performance was assessed using accuracy and macro F1-score metrics. The rainfall classes used in this study consist of four categories: no rain, light rain, moderate rain, and heavy rain. The results indicate that outlier removal improves the accuracy of all evaluated algorithms, with the most substantial improvement observed in the Decision Tree model with accuracy improved from 45.71% to 57.36% and macro F1-score from 28.99% to 38.78%. Overall, the implementation of outlier removal yields more balanced and representative rainfall classification results, potentially serving as a basis for future quantitative rainfall regression studies.]]>
Copyrights © 2026
