Prediction of road traffic accident severity based on multi-model fusion

Wenting Lu; Mingxia Huang; Rongze Yu

doi:10.1117/12.2689859

7 September 2023 Prediction of road traffic accident severity based on multi-model fusion

Wenting Lu, Mingxia Huang, Rongze Yu

Proceedings Volume 12790, Eighth International Conference on Electromechanical Control Technology and Transportation (ICECTT 2023); 127906K (2023) https://doi.org/10.1117/12.2689859
Event: 8th International Conference on Electromechanical Control Technology and Transportation (ICECTT 2023), 2023, Hangzhou, China

Abstract

In recent years, the prevention of road traffic accidents has emerged as a focal point in traffic safety research. This study introduces a multimodal fusion-based approach to predict the severity of road traffic accidents. Initially, the SMOTE(Synthetic Minority Over-sampling Technique) Tomek method is employed to address imbalanced data, and a multimodal weighted fusion model based on Voting principles is established. Logistic regression, SVM(Support Vector Machine), and decision trees are chosen as base models, with the weighted average F1 values of each base model serving as weights to create the fusion model. Subsequently, the SHAP (SHapley Additive exPlanation) model is incorporated to interpret the analysis results of the fusion model and compare the relationships between the influence of each feature variable on the dependent variable. Finally, by comparing the evaluation metrics of overall data, urban data, and rural data within the fusion model, it is found that the training scores of all three data types exceed 0.90. This indicates that the model fusion achieves a higher accuracy rate and F1 value in predicting road accident severity compared to a single classifier. Moreover, the evaluation metrics of the model fusion are generally superior to those of the three classifiers, with an accuracy increase of 0.01 percentage points compared to SVM, and improvements in both precision and recall. Thus, the multimodal fusion algorithm proposed in this study demonstrates higher evaluation metrics.

Citation Download Citation

Wenting Lu, Mingxia Huang, and Rongze Yu "Prediction of road traffic accident severity based on multi-model fusion", Proc. SPIE 12790, Eighth International Conference on Electromechanical Control Technology and Transportation (ICECTT 2023), 127906K (7 September 2023); https://doi.org/10.1117/12.2689859

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
5 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Data modeling

Roads

Data fusion

Decision trees

Education and training

Visual process modeling

Statistical analysis

Show All Keywords

Keywords/Phrases

Search In:

Publication Years