Development of a Hybrid Model for Spatio-Temporal Prediction and Analysis of Road Traffic Accidents under Humid Climatic Conditions (Case Study: Rasht City)

Development of a Hybrid Model for Spatio-Temporal Prediction and Analysis of Road Traffic Accidents under Humid Climatic Conditions (Case Study: Rasht City)

Ali khorram Soleymandarabi1 alireza azizi kolachahi2

1) M.SC highway and transportaion engineer
2) M.SC transportaion engineer

IOI Code : AACA-ABEGHD
Publication : 10th.International Conference on Researches in Science & Engineering & 7th.International Congress on Civil, Architecture and Urbanism in Asia(10icrsie.com)
Abstract :
    Road traffic accidents represent a major urban safety challenge, particularly in cities with humid climatic conditions. Factors such as frequent rainfall, high humidity, and reduced pavement friction significantly influence the spatial and temporal patterns of traffic accidents. This study aims to develop a hybrid machine learning model for predicting and analyzing the spatiotemporal distribution of traffic accidents under humid climatic conditions, with a case study of Rasht city, Iran. The research utilizes real traffic accident data recorded in 2024 along with rainfall data as auxiliary variables. After data integration and preprocessing, spatiotemporal analyses were conducted to identify accident hotspots and temporal patterns. Subsequently, a hybrid model combining Random Forest and XGBoost algorithms was developed and evaluated against baseline models. The results indicate that integrating climatic variables with spatial and temporal features significantly improves prediction accuracy and enhances the identification of high-risk locations during rainy conditions. The proposed approach provides valuable insights for traffic safety management and supports data-driven decision-making aimed at reducing traffic accidents in humid urban environments.Keywords:Traffic accidents, Humid climate, Machine learning, Spatiotemporal analysis, Rasht city

Keywords : road traffic accidents humid climate hybrid development spatio temporal analysis rasht city