Covariate Selection from Telematics Car Driving Data

18 Pages Posted: 19 Dec 2016  

Mario V. Wuthrich

RiskLab, ETH Zurich

Date Written: December 19, 2016

Abstract

Car insurance companies have started to collect high-frequency GPS location data of their car drivers. This data provides detailed information about the driving habits and driving styles of individual car drivers. We illustrate how this data can be analyzed using techniques from pattern recognition and machine learning. In particular, we describe how driving styles can be categorized so that they can be used for a regression analysis in car insurance pricing.

Keywords: telematics data, driving habits, driving styles, regression, categorical classes, pattern recognition, clustering, K-means clustering, unsupervised learning, machine learning

JEL Classification: G22, G28

Suggested Citation

Wuthrich, Mario V., Covariate Selection from Telematics Car Driving Data (December 19, 2016). Available at SSRN: https://ssrn.com/abstract=2887357 or http://dx.doi.org/10.2139/ssrn.2887357

Mario V. Wuthrich (Contact Author)

RiskLab, ETH Zurich ( email )

Department of Mathematics
Ramistrasse 101
Zurich, 8092
Switzerland

Paper statistics

Downloads
168
Rank
144,021
Abstract Views
366