<![CDATA[Most machine learning algorithms tend to yield optimal results when trained on datasets with balanced class proportions. However, their performance usually declines when applied to data with significant class imbalance. To address this issue, this study utilizes the Synthetic Minority Oversampling Technique (SMOTE) to improve class distribution before model training. Several classification algorithms were employed, including Decision Tree, K-Nearest Neighbors, Logistic Regression, Support Vector Machine, and Random Forest. Experimental results reveal that the Random Forest model produced the highest accuracy (95.70%) and the best F1-score, demonstrating a well-balanced trade-off between precision and recall. In contrast, the Logistic Regression algorithm achieved the highest recall (74.20%), indicating better sensitivity in identifying positive instances despite a lower F1-score. These outcomes highlight the importance of choosing appropriate classification methods based on the specific evaluation goals whether prioritizing accuracy, recall, or overall model balance.]]>
Copyrights © 2025
