Enhancing Emergency Response in Road Accidents: a severity Prediction Frame work Using RF-RFE and Deep Learning Model

Abstract

Road accidents remain a major public safety concern, where timely and accurate assessment of accident severity is critical for improving emergency response and reducing fatalities. This study proposes a severity prediction framework that integrates Random Forest–Recursive Feature Elimination (RF-RFE) with a deep learning model to enhance prediction accuracy and decision-making. The RF-RFE technique is employed to identify and select the most relevant features from large and complex accident datasets, effectively reducing dimensionality and improving model efficiency. The selected features are then fed into a deep learning model capable of capturing complex nonlinear relationships among variables to predict accident severity levels. The proposed framework aims to assist emergency services in prioritizing responses, optimizing resource allocation, and minimizing response time. Experimental results demonstrate that the hybrid approach outperforms traditional machine learning models in terms of accuracy, precision, and recall. Overall, this system provides a robust and scalable solution for intelligent traffic management and emergency response enhancement, contributing to safer road environments.