<![CDATA[Vehicle insurance companies increasingly rely on data-driven marketing strategies to identify prospective customers who are likely to respond positively to insurance offers. However, customer response prediction is challenging due to class imbalance, where non-responsive customers substantially outnumber responsive ones. This study aims to compare the performance of Logistic Regression and Random Forest models in predicting customer responses to vehicle insurance products using the Synthetic Minority Oversampling Technique (SMOTE). The analysis was conducted using the Vehicle Insurance dataset obtained from Kaggle. Experimental results indicate that Random Forest achieved the best overall performance, with an accuracy of 0.80, a positive-class F1-score of 0.59, and a ROC–AUC score of 0.88. In contrast, Logistic Regression produced a higher positive-class recall of 0.98 but a lower precision of 0.35, indicating a greater tendency to generate false-positive predictions. Feature importance analysis revealed that Previously_Insured, Vehicle_Damage, and Age were the most influential factors affecting customer responses. These findings suggest that the combination of Random Forest and SMOTE provides an effective approach for handling imbalanced data and improving customer response prediction in vehicle insurance marketing campaigns.]]>
Copyrights © 2026
