Predicting and Explaining Traffic Crash Severity Through Crash Feature Selection

2025-08-15

AI Summary

This study addresses traffic accident severity prediction and causal attribution by constructing a large-scale dataset comprising over 2.3 million vehicle crashes. We propose a transparent, reproducible hierarchical analytical framework that integrates JADBio AutoML for automated model selection, SHAP-based global and local interpretability analysis, and ridge logistic regression to enhance classification robustness. A key contribution is the identification of environmental and contextual factors—such as weather conditions, time of day, and road type—as significantly more predictive of accident severity than conventional risk indicators (e.g., drunk driving). The final model achieves an AUC-ROC of 84.9% on the held-out test set and consistently identifies 17 critical predictors spanning demographic, environmental, vehicular, and behavioral dimensions. These findings provide actionable, data-driven insights for precision traffic safety governance and evidence-based policy formulation.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Natural Language Processing: Safety and Robustness

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Motor vehicle crashes remain a leading cause of injury and death worldwide, necessitating data-driven approaches to understand and mitigate crash severity. This study introduces a curated dataset of more than 3 million people involved in accidents in Ohio over six years (2017-2022), aggregated to more than 2.3 million vehicle-level records for predictive analysis. The primary contribution is a transparent and reproducible methodology that combines Automated Machine Learning (AutoML) and explainable artificial intelligence (AI) to identify and interpret key risk factors associated with severe crashes. Using the JADBio AutoML platform, predictive models were constructed to distinguish between severe and non-severe crash outcomes. The models underwent rigorous feature selection across stratified training subsets, and their outputs were interpreted using SHapley Additive exPlanations (SHAP) to quantify the contribution of individual features. A final Ridge Logistic Regression model achieved an AUC-ROC of 85.6% on the training set and 84.9% on a hold-out test set, with 17 features consistently identified as the most influential predictors. Key features spanned demographic, environmental, vehicle, human, and operational categories, including location type, posted speed, minimum occupant age, and pre-crash action. Notably, certain traditionally emphasized factors, such as alcohol or drug impairment, were less influential in the final model compared to environmental and contextual variables. Emphasizing methodological rigor and interpretability over mere predictive performance, this study offers a scalable framework to support Vision Zero with aligned interventions and advanced data-informed traffic safety policy.

Problem

• Predict traffic crash severity using data-driven methods
• Identify key risk factors with explainable AI techniques
• Improve traffic safety policies through interpretable predictive models

Innovation

• AutoML for crash severity prediction
• SHAP for feature importance interpretation
• Ridge Logistic Regression with high AUC