Applying Tabular Deep Learning Models to Estimate Crash Injury Types of Young Motorcyclists

2025-03-13

AI Summary

This study addresses the challenge of predicting injury severity (fatal/severe, moderate/minor, none) among 10,726 motorcycle crashes involving riders aged 15–24 in Texas (2017–2022). To tackle severe class imbalance in this three-class problem, we propose an end-to-end tabular deep learning framework integrating ARMNet and MambaNet—novel architectures introduced here for traffic injury prediction—combined with SMOTE-Tomek Link resampling and a feature embedding–attention fusion module. Our model achieves F1-scores >0.90 for fatal/severe and no-injury classes; ARMNet attains 87% overall accuracy and significantly improves robustness in moderate/minor injury classification. Furthermore, interpretability analysis reveals distinct behavioral (e.g., speeding, helmet non-use), environmental, and demographic determinants of injury outcomes among youth, uncovering differential impact mechanisms. These findings provide quantifiable, interpretable evidence to inform targeted helmet enforcement and age-stratified safety education policies.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Computer Vision: Interpretability, Explainability, and Transparency
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Young motorcyclists, particularly those aged 15 to 24 years old, face a heightened risk of severe crashes due to factors such as speeding, traffic violations, and helmet usage. This study aims to identify key factors influencing crash severity by analyzing 10,726 young motorcyclist crashes in Texas from 2017 to 2022. Two advanced tabular deep learning models, ARMNet and MambaNet, were employed, using an advanced resampling technique to address class imbalance. The models were trained to classify crashes into three severity levels, Fatal or Severe, Moderate or Minor, and No Injury. ARMNet achieved an accuracy of 87 percent, outperforming 86 percent of Mambanet, with both models excelling in predicting severe and no injury crashes while facing challenges in moderate crash classification. Key findings highlight the significant influence of demographic, environmental, and behavioral factors on crash outcomes. The study underscores the need for targeted interventions, including stricter helmet enforcement and educational programs customized to young motorcyclists. These insights provide valuable guidance for policymakers in developing evidence-based strategies to enhance motorcyclist safety and reduce crash severity.

Problem

• Identify factors influencing crash severity in young motorcyclists.
• Classify crash severity using tabular deep learning models.
• Provide insights for targeted interventions to reduce crash severity.

Innovation

• ARMNet and MambaNet tabular deep learning models
• Advanced resampling for class imbalance handling
• Classification of crash severity into three levels