Skip to main content
Open Access Publications from the University of California


UCLA Electronic Theses and Dissertations bannerUCLA

Severity Prediction and Time-Series Analysis of Vehicle Accidents Using Statistical Models


This study explores factors that effect vehicle accidents, predicts the severity of accidents through logistic regression, and forecasts the number of future accidents to occur using time-series analysis. From insights gathered during exploration, a final dataset is prepared for the use of a logistic regression model. The final model predicts whether or not an accident will be severe with an accuracy of 82%, and reveals the three main features that statistically contribute to the odds of an accident having a severe impact on traffic. Finally, a time-series analysis is run in order to model the number of accidents that can occur on a given day using historical data. This paper evaluates the dataset in ways that have yet to be explored, and provides a great baseline understanding of what is possible for the future of transportation.

Main Content
For improved accessibility of PDF content, download the file to your device.
Current View