×

Improving automobile insurance claims frequency prediction with telematics car driving data. (English) Zbl 07540865

Summary: Novel navigation applications provide a driving behavior score for each finished trip to promote safe driving, which is mainly based on experts’ domain knowledge. In this paper, with automobile insurance claims data and associated telematics car driving data, we propose a supervised driving risk scoring neural network model. This one-dimensional convolutional neural network takes time series of individual car driving trips as input and returns a risk score in the unit range of (0,1). By incorporating credibility average risk score of each driver, the classical Poisson generalized linear model for automobile insurance claims frequency prediction can be improved significantly. Hence, compared with non-telematics-based insurers, telematics-based insurers can discover more heterogeneity in their portfolio and attract safer drivers with premiums discounts.

MSC:

91G05 Actuarial mathematics
PDF BibTeX XML Cite
Full Text: DOI

References:

[1] Ayuso, M., Guillen, M. and Pérez-Marín, A.M. (2016a) Telematics and gender discrimination: Some usage-based evidence on whether men’s risk of accidents differs from women’s. Risks, 4(2), 10.
[2] Ayuso, M., Guillen, M. and Pérez-Marín, A.M. (2016b) Using GPS data to analyse the distance traveled to the first accident at fault in pay-as-you-drive insurance. Transportation Research Part C: Emerging Technologies, 68, 160-167.
[3] Boucher, J.-P., Côté, S. and Guillen, M. (2017) Exposure as duration and distance in telematics motor insurance using generalized additive models. Risks, 5(4): 54.
[4] Carvalho, E., Ferreira, B.V., Ferreira, J., De Souza, C., Carvalho, H.V., Suhara, Y., Pentland, A.S. and Pessin, G. (2017) Exploiting the use of recurrent neural networks for driver behavior profiling. In 2017 International Joint Conference on Neural Networks (IJCNN), pp. 3016-3021.
[5] Cather, D. A. (2020) Reconsidering insurance discrimination and adverse selection in an era of data analytics. Geneva Papers on Risk and Insurance - Issues and Practice, 45, 426-456.
[6] Cevolini, A. and Esposito, E. (2020) From pool to profile: Social consequences of algorithmic prediction in insurance. Big Data & Society, 7(2).
[7] Delong, L., Lindholm, M. and Wüthrich, M.V. (2021) Making Tweedie’s compound Poisson model more accessible. European Actuarial Journal, 11, 185-226. · Zbl 1485.91208
[8] Denuit, M., Guillen, M. and Trufin, J. (2019) Multivariate credibility modelling for usage-based motor insurance pricing with behavioural data. Annals of Actuarial Science, 13(2), 378-399.
[9] Gao, G., Shi, Y. and Wang, H. (2021) Telematics car driving data analytics. The Society of Actuaries Research Institute, pp. 1-50.
[10] Gao, G., Wang, H. and Wüthrich, M.V. (2021) Boosting poisson regression models with telematics car driving data. Machine Learning. · Zbl 07510312
[11] Geyer, A., Kremslehner, D. and Muermann, A. (2020) Asymmetric information in automobile insurance: Evidence from driving behavior. Journal of Risk and Insurance, 87(4), 969-995.
[12] Girma, A., Yan, X. and Homaifar, A. (2019) Driver identification based on vehicle telematics data using LSTM recurrent neural network. In 2019 IEEE 31st International Conference on Tools with Artificial Intelligence (ICTAI), pp. 894-902.
[13] Guillen, M., Nielsen, J.P., Pérez-Marín, A.M. and Elpidorou, V. (2020) Can automobile insurance telematics predict the risk of near-miss events?North American Actuarial Journal, 24(1), 141-152. · Zbl 1437.91392
[14] Ho, S.-H., Wong, Y.D. and Chang, V.W.C. (2014) Developing Singapore driving cycle for passenger cars to estimate fuel consumption and vehicular emissions. Atmospheric Environment, 97, 353-362.
[15] Hornik, K., Stinchcombe, M. and White, H. (1989) Multilayer feedforward networks are universal approximators, Neural Networks, 2(5), 359-366. · Zbl 1383.92015
[16] Hu, X., Zhu, X., Ma, Y.-L., Chiu, Y.-C. and Tang, Q. (2019) Advancing usage-based insurance - A contextual driving risk modelling and analysis approach. IET Intelligent Transport Systems, 13(3), 453-460.
[17] Huang, Y. and Meng, S. (2019). Automobile insurance classification ratemaking based on telematics driving data, Decision Support Systems, 127.
[18] Hung, W. T., Tong, H. Y., Lee, C. P., Ha, K. and Pao, L. Y. (2007). Development of practical driving cycle construction methodology: A case study in Hong Kong. Transportation Research Part D: Transport and Environment, 12(2), 115-128.
[19] Joubert, J.W., De Beer, D. and De Koker, N. (2016) Combining accelerometer data and contextual variables to evaluate the risk of driver behaviour. Transportation Research Part F: Traffic Psychology and Behaviour, 4, 80-96.
[20] Kamble, S.H., Mathew, T.V. and Sharma, G.K. (2009) Development of real-world driving cycle: Case study of Pune, India. Transportation Research Part D: Transport and Environment, 14(2), 132-140.
[21] Lecun, Y., Bengio, Y. and Hinton, G. (n.d.) Deep learning. Nature, 521, 436-444.
[22] Lemaire, J., Park, S. C. and Wang, K. (2016) The use of annual mileage as a rating variable. ASTIN Bulletin, 46(1), 39-69. · Zbl 1390.62213
[23] Ma, Y.-L., Zhu, X., Hu, X. and Chiu, Y.-C. (2018) The use of context-sensitive insurance telematics data in auto insurance rate making. Transportation Research Part A: Policy and Practice, 113, 243-258.
[24] . (2019) Privacy: A growing risk in the insurance industry. Creative Components, 410.
[25] Masood, S., Rai, A., Aggarwal, A., Doja, M.N. and Ahmad, M. (2020) Detecting distraction of drivers using convolutional neural network. Pattern Recognition Letters, 139, 79-85.
[26] Paefgen, J., Staake, T. and Fleisch, E. (2014) Multivariate exposure modeling of accident risk: Insights from pay-as-you-drive insurance data. Transportation Research Part A: Policy and Practice, 61, 27-40.
[27] Richman, R. (2020) AI in actuarial science - A review of recent advances. Annals of Actuarial Science.
[28] Savelonas, M., Mantzekis, D., Labiris, N., Tsakiri, A., Karkanis, S. and Spyrou, E. (2020) Hybrid time-series representation for the classification of driving behaviour. In 2020 15th International Workshop on Semantic and Social Media Adaptation and Personalization (SMA), pp. 1-6.
[29] Singh, D. (2016) Using convolutional neural networks to perform classification on state farm insurance driver images, Technical report, Stanford University.
[30] Verbelen, R., Antonio, K. and Claeskens, G. (2018) Unraveling the predictive power of telematics data in car insurance pricing. Journal of the Royal Statistical Society: Series C (Applied Statistics), 67(5), 1275-1304.
[31] Wahlström, J., Skog, I. and Händel, P. (2015) Detection of dangerous cornering in GNSS-data-driven insurance telematics. IEEE Transactions on Intelligent Transportation Systems, 16(6), 3073-3083.
[32] Wahlström, J., Skog, I. and Händel, P. (2017) Smartphone-based vehicle telematics: A ten-year anniversary. IEEE Transactions on Intelligent Transportation Systems,18(10), 2802-2825.
[33] Wahlström, J., Skog, I., Nordström, R.L. and Händel, P. (2018) Fusion of OBD and GNSS measurements of speed. IEEE Transactions on Instrumentation and Measurement67(7), 1659-1667.
[34] Weidner, W., Transchel, F.W.G. and Weidner, R. (2016) Classification of scale-sensitive telematic observables for riskindividual pricing. European Actuarial Journal, 6(1), 3-24. · Zbl 1415.91167
[35] Weidner, W., Transchel, F.W.G. and Weidner, R. (2017) Telematic driving profile classification in car insurance pricing. Annals of Actuarial Science, 11(2), 213-236.
[36] Wüthrich, M. V. (2021) The balance property in neural network modelling. Statistical Theory and Related Fields, 1-9.
[37] Yang, Y., Qian, W. and Zou, H. (2016) Insurance premium prediction via gradient tree-boosted Tweedie compound Poisson models. Journal of Business and Economic Statistics, 36(3), 456-470.
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.