Date of Award
2024
Degree Type
Thesis
Degree Name
Master of Science in Statistics
Department
Computer Science and Statistics
First Advisor
Guangyu Zhu
Abstract
The report by the Insurance Information Institute underscores a notable increase in both claims frequency and severity, particularly highlighting a significant surge in accident claim severity within US car insurance from 2010 to 2019. Simultaneously, there has been a marked rise in the average expenditure on US car insurance during this timeframe. These shifts emphasize the critical need for accurate predictions to fine-tune premium adjustments and enhance the accessibility of car insurance coverage for a broader demographic of drivers. Consequently, numerous insurance companies are transitioning from traditional methodologies to incorporate machine learning (ML) techniques, providing a more sophisticated and reliable framework for generating outcomes. Nonetheless, the challenge persists in selecting the most optimal ML predictive model to effectively identify probable claims or potential premium defaulters. This study tackles these complexities by employing diverse classification methods and proposing specific techniques for feature selection and data resampling, with the overarching goal of constructing comprehensive classification models tailored for in-depth claim analysis.
Recommended Citation
Roy, Sankhadip, "MOTOR INSURANCE CLAIMS PREDICTION: COMPARATIVE STUDY USING MACHINE LEARNING" (2024). Open Access Master's Theses. Paper 2501.
https://digitalcommons.uri.edu/theses/2501