Deep Learning, Machine Learning, or Statistical Models for Weather-Related Crash Severity Prediction [Research Brief]
-
2023-12-01
-
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Research Brief
-
Corporate Publisher:
-
Abstract:In the realm of traffic safety, weather-related road traffic crashes pose a significant public health concern, leading to numerous injuries and fatalities. Past research has explored crash severity prediction using statistical machine learning (ML) and deep learning (DL)models, each with its strengths and limitations. This study strategically selected a diverse set of models, including ordered logit and ordered probit models (OLM and OPM), random forest (RF),XGBoost, multi-layer perceptron neural network(MLP), and TabNet, to comprehensively assess their effectiveness in predicting crash severity while considering weather conditions. By including this array of models, this research offers valuable insights for traffic safety professionals to predict crash severity levels, enabling them to make informed decisions and allocate resources effectively to reduce the impact of weather-related crashes on road safety.
-
Format:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type: