Nightly Automobile Claims Prediction from Telematics-Derived Features: A Multilevel Approach
In recent years it has become possible to collect GPS data from drivers and to incorporate these data into automobile insurance pricing for the driver. These data are continuously collected and processed nightly into metadata consisting of mileage and time summaries of each discrete trip taken, and...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-06-01
|
| Series: | Risks |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-9091/10/6/118 |
| _version_ | 1797482621384523776 |
|---|---|
| author | Allen R. Williams Yoolim Jin Anthony Duer Tuka Alhani Mohammad Ghassemi |
| author_facet | Allen R. Williams Yoolim Jin Anthony Duer Tuka Alhani Mohammad Ghassemi |
| author_sort | Allen R. Williams |
| collection | DOAJ |
| description | In recent years it has become possible to collect GPS data from drivers and to incorporate these data into automobile insurance pricing for the driver. These data are continuously collected and processed nightly into metadata consisting of mileage and time summaries of each discrete trip taken, and a set of behavioral scores describing attributes of the trip (e.g, driver fatigue or driver distraction), so we examine whether it can be used to identify periods of increased risk by successfully classifying trips that occur immediately before a trip in which there was an incident leading to a claim for that driver. Identification of periods of increased risk for a driver is valuable because it creates an opportunity for intervention and, potentially, avoidance of a claim. We examine metadata for each trip a driver takes and train a classifier to predict whether the following trip is one in which a claim occurs for that driver. By achieving an area under the receiver–operator characteristic above 0.6, we show that it is possible to predict claims in advance. Additionally, we compare the predictive power, as measured by the area under the receiver–operator characteristic of XGBoost classifiers trained to predict whether a driver will have a claim using exposure features such as driven miles, and those trained using behavioral features such as a computed speed score. |
| first_indexed | 2024-03-09T22:35:00Z |
| format | Article |
| id | doaj.art-a142d0ec91a345458d042698f91e103d |
| institution | Directory Open Access Journal |
| issn | 2227-9091 |
| language | English |
| last_indexed | 2024-03-09T22:35:00Z |
| publishDate | 2022-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Risks |
| spelling | doaj.art-a142d0ec91a345458d042698f91e103d2023-11-23T18:50:05ZengMDPI AGRisks2227-90912022-06-0110611810.3390/risks10060118Nightly Automobile Claims Prediction from Telematics-Derived Features: A Multilevel ApproachAllen R. Williams0Yoolim Jin1Anthony Duer2Tuka Alhani3Mohammad Ghassemi4Department of Computer Science & Engineering, Michigan State University, East Lansing, MI 48842, USACSAA Insurance Group 3055 Oak Road, Walnut Creek, CA 94597, USACSAA Insurance Group 3055 Oak Road, Walnut Creek, CA 94597, USAEngineering Division, New York University Abu Dhabi, Saadiyat Campus, Abu Dhabi P.O. Box 129188, United Arab EmiratesDepartment of Computer Science & Engineering, Michigan State University, East Lansing, MI 48842, USAIn recent years it has become possible to collect GPS data from drivers and to incorporate these data into automobile insurance pricing for the driver. These data are continuously collected and processed nightly into metadata consisting of mileage and time summaries of each discrete trip taken, and a set of behavioral scores describing attributes of the trip (e.g, driver fatigue or driver distraction), so we examine whether it can be used to identify periods of increased risk by successfully classifying trips that occur immediately before a trip in which there was an incident leading to a claim for that driver. Identification of periods of increased risk for a driver is valuable because it creates an opportunity for intervention and, potentially, avoidance of a claim. We examine metadata for each trip a driver takes and train a classifier to predict whether the following trip is one in which a claim occurs for that driver. By achieving an area under the receiver–operator characteristic above 0.6, we show that it is possible to predict claims in advance. Additionally, we compare the predictive power, as measured by the area under the receiver–operator characteristic of XGBoost classifiers trained to predict whether a driver will have a claim using exposure features such as driven miles, and those trained using behavioral features such as a computed speed score.https://www.mdpi.com/2227-9091/10/6/118telematicsusage based insurancerisk mitigation |
| spellingShingle | Allen R. Williams Yoolim Jin Anthony Duer Tuka Alhani Mohammad Ghassemi Nightly Automobile Claims Prediction from Telematics-Derived Features: A Multilevel Approach Risks telematics usage based insurance risk mitigation |
| title | Nightly Automobile Claims Prediction from Telematics-Derived Features: A Multilevel Approach |
| title_full | Nightly Automobile Claims Prediction from Telematics-Derived Features: A Multilevel Approach |
| title_fullStr | Nightly Automobile Claims Prediction from Telematics-Derived Features: A Multilevel Approach |
| title_full_unstemmed | Nightly Automobile Claims Prediction from Telematics-Derived Features: A Multilevel Approach |
| title_short | Nightly Automobile Claims Prediction from Telematics-Derived Features: A Multilevel Approach |
| title_sort | nightly automobile claims prediction from telematics derived features a multilevel approach |
| topic | telematics usage based insurance risk mitigation |
| url | https://www.mdpi.com/2227-9091/10/6/118 |
| work_keys_str_mv | AT allenrwilliams nightlyautomobileclaimspredictionfromtelematicsderivedfeaturesamultilevelapproach AT yoolimjin nightlyautomobileclaimspredictionfromtelematicsderivedfeaturesamultilevelapproach AT anthonyduer nightlyautomobileclaimspredictionfromtelematicsderivedfeaturesamultilevelapproach AT tukaalhani nightlyautomobileclaimspredictionfromtelematicsderivedfeaturesamultilevelapproach AT mohammadghassemi nightlyautomobileclaimspredictionfromtelematicsderivedfeaturesamultilevelapproach |