Related Experiment Video
Updated: Nov 24, 2025
07:31
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
7.4K
Deep-Learning-Based Prediction of High-Risk Taxi Drivers Using Wellness Data.
Seolyoung Lee1, Jae Hun Kim1, Jiwon Park2
1Research Institute of Engineering Technology, Hanyang University Erica Campus, Ansan 15588, Korea.
International Journal of Environmental Research and Public Health
|December 23, 2020
Summary
This study uses deep learning to predict high-risk taxi drivers by analyzing driver wellness factors. The best model achieved 86% accuracy, aiding in developing taxi driver management and safety systems.
Area of Science:
- Transportation safety
- Human factors in driving
- Machine learning applications
Background:
- Driver wellness is crucial for identifying high-risk taxi drivers.
- Personal characteristics influence driving behavior and accident risk.
- Understanding wellness factors can improve road safety.
Purpose of the Study:
- To predict high-risk taxi drivers using a deep learning model.
- To identify key driver wellness factors impacting accident severity.
- To develop a data-driven approach for taxi driver risk assessment.
Main Methods:
- In-depth interviews collected taxi driver wellness data.
- Random forest models prioritized factors influencing accident severity.
- Artificial neural network models were optimized for various input scenarios.
Main Results:
- The top-performing model included variables up to the 16th priority.
- This model achieved 86% classification accuracy.
- An F1-score of 0.77 was recorded for the best predictive model.
Conclusions:
- A wellness-based deep learning model can predict high-risk taxi drivers.
- The model supports the development of taxi driver management systems.
- Findings can inform customized traffic safety measures for commercial drivers.
