PERBANDINGAN PREDIKSI KEBERLANJUTAN POLIS ASURANSI DENGAN MENGGUNAKAN ALGORITMA NAIVE BAYES, KNN, DAN SVM

Abstract

This research investigates insurance policy continuation using Naive Bayes, K-Nearest Neighbors (KNN), and Support Vector Machine (SVM) algorithms. Data from PT Asuransi ABC, consisting of 35,420 data points, was analyzed to predict policy cancellations. The findings suggest that KNN outperforms other algorithms in handling non-linear data. KNN successfully identified reasons for policy cancellations with greater accuracy, while Gaussian Naive Bayes and SVM require further adjustments to enhance their accuracy. This research provides valuable insights for insurance companies in developing strategies to increase customer retention and reduce policy cancellation rates. This is achieved by understanding policy cancellation patterns, enabling companies to take proactive steps to provide better services and tailor their insurance products to meet customer needs.