Abstract
In recent years, technological improvements have provided a variety of new opportunities for insurance companies to adopt telematics devices in line with usage-based insurance models. This paper sheds new light on the application of big data analytics for car insurance companies that may help to estimate the risks associated with individual policyholders based on complex driving patterns. We propose a conceptual framework that describes the structural design of a risk predictor model for insurance customers and combines the value of telematics data with deep learning algorithms. The model’s components consist of data transformation, criteria mining, risk modelling, driving style detection, and risk prediction. The expected outcome is our methodology that generates more accurate results than other methods in this area.
Recommended Citation
Siami, Mohammad; Naderpour, Mohsen; and Lu, Jie, "Generating a Risk Profile for Car Insurance Policyholders: a Deep Learning Conceptual Model" (2017). ACIS 2017 Proceedings. 30.
https://aisel.aisnet.org/acis2017/30