International Conference on Computer and Knowledge Engineering

Home / 14th International Conference on Computer and Knowledge Engineering

A Cost-Sensitive Genetic Algorithm for Customer Segmentation in Auto Insurances

Authors :

Alireza Khajenoori¹ Mohammad Saniee Abadeh² Mohsen Mohammadzadeh³

1- Faculty of Interdisciplinary Science and Technology Tarbiat Modares University 2- Faculty of Electrical and Computer Engineering Tarbiat Modares University 3- Faculty of Interdisciplinary Science and Technology Tarbiat Modares University

Keywords :

Auto insurance،Customer segmentation،Cost-sensitive learning

Abstract :

In the auto insurance industry, accurate risk assessment is essential for determining fair premiums and managing claims effectively. As customer data grows in volume and complexity, machine learning has become crucial for precise customer segmentation and risk prediction. However, the industry faces significant challenges, particularly the class imbalance in data and the unequal costs associated with misclassification errors, where some errors are far more costly than others. This study compares two approaches to address these issues: traditional sampling techniques commonly used to mitigate class imbalance, and a cost-sensitive learning framework that optimizes the cost matrix using a genetic algorithm to minimize the financial impact of misclassification. The findings demonstrate that the cost-sensitive approach significantly enhances overall model performance by more effectively prioritizing costly errors, leading to more accurate and economically sound decision-making. This research highlights the importance of integrating advanced machine learning techniques in the insurance sector, showing that such approaches can substantially improve the fairness and efficiency of risk prediction models, ultimately benefiting insurers by enabling more precise premium setting and increasing customer satisfaction.

Papers List

List of archived papers

LightFedSelect: A Lightweight Framework for Byzantine-Robust Federated Learning

Seyed Saeed Razavi - Seyed Arsalan Vasegh Rahim Parvar - Soroosh Dadashi Pakdeh - Mohammad Matin Rezaeifard - Morteza Mollaie Chafi - Reza Ebrahimi Atani