Related Experiment Video
Updated: Sep 30, 2025
11:12
Driving Simulation in the Clinic: Testing Visual Exploratory Behavior in Daily Life Activities in Patients with Visual Field Defects
Published on: September 18, 2012
17.5K
Predicting and Analyzing Road Traffic Injury Severity Using Boosting-Based Ensemble Learning Models with SHAPley
Sheng Dong1, Afaq Khattak2, Irfan Ullah3
1School of Civil and Transportation Engineering, Ningbo University of Technology, Fenghua Road No. 201, Ningbo 315211, China.
Summary
Predicting road traffic injury severity is crucial. This study introduces interpretable machine learning models, with LightGBM and SHAP analysis identifying key risk factors like driver age and accident cause for enhanced safety.
Area of Science:
- Machine Learning and Artificial Intelligence
- Transportation Safety
- Public Health
Background:
- Road traffic accidents pose a significant global threat, causing numerous fatalities, injuries, and economic losses annually.
- Existing ensemble learning models for injury severity prediction often function as "black boxes," limiting their practical interpretability.
- Understanding contributing factors to road traffic injury severity is vital for developing effective mitigation strategies.
Purpose of the Study:
- To develop interpretable predictive models for road traffic injury severity using boosting-based ensemble learning.
- To identify and rank the most significant risk variables contributing to injury severity.
- To provide insights into the factors influencing fatal injuries in road traffic accidents.
Main Methods:
- Proposed four boosting-based ensemble learning models: Natural Gradient Boosting, Adaptive Gradient Boosting, Categorical Gradient Boosting, and Light Gradient Boosting Machine (LightGBM).
- Utilized SHapley Additive exPlanations (SHAP) analysis to interpret the optimal predictive model and rank risk variables.
- Applied the models to accident data from Pakistan's National Highway N-5 (2015-2019).
Main Results:
- LightGBM demonstrated superior performance, achieving the highest accuracy (73.63%), precision (72.61%), recall (70.09%), F1-score (70.81%), and AUC (0.71).
- SHAP analysis identified Month_of_Year, Cause_of_Accident, Driver_Age, and Collision_Type as significant predictors of injury severity.
- Key risk factors for fatal injuries include young drivers, collisions involving trailers and pedestrians, and rear-end collisions.
Conclusions:
- The combination of LightGBM and SHAP provides an effective approach for developing interpretable models for road traffic injury severity prediction.
- Identifying specific high-risk scenarios, such as young drivers involved in trailer-related accidents, can inform targeted safety interventions.
- The findings offer valuable insights for policymakers and safety engineers to mitigate the risks associated with road traffic accidents.
Related Concept Videos
01:08Determination of Expected Frequency
2.3K
Suppose one wants to test independence between the two variables of a contingency table. The values in the table constitute the observed frequencies of the dataset. But how does one determine the expected frequency of the dataset? One of the important assumptions is that the two variables are independent, which means the variables do not influence each other. For independent variables, the statistical probability of any event involving both variables is calculated by multiplying the individual...
2.3K
01:23Mechanistic Models: Compartment Models in Individual and Population Analysis
92
Mechanistic models are utilized in individual analysis using single-source data, but imperfections arise due to data collection errors, preventing perfect prediction of observed data. The mathematical equation involves known values (Xi), observed concentrations (Ci), measurement errors (εi), model parameters (ϕj), and the related function (ƒi) for i number of values. Different least-squares metrics quantify differences between predicted and observed values. The ordinary least...
92