OPTIMIZING EMERGENCY RESPONSE THROUGH DEEP LEARNING-BASED AMBULANCE POSITIONING
Keywords:
Road Accidents , Ambulance Prepositioning , Deep Learning , Healthcare Services , Optimal Location Prediction , Cat2Vec, K-means , GMMAbstract
Traffic accidents continue to be a significant global issue as a result of the significant number of injuries and fatalities they result in. In these instances, prompt medical assistance is essential, and the utilisation of ambulances that are prepositioned rather than dispatched on demand can significantly reduce response times and enhance the quality of the outcome. This Paper delineates a deep-embedded clustering-based approach to ambulance placement prediction. The model takes into account the geographic patterns and characteristics that affect the frequency of traffic incidents in order to enhance the accuracy of its forecasts. Cat2Vec, a deep learning-based method, maintains spatial and contextual patterns while training models for real-time performance. The framework is contrasted with traditional clustering methods, including Agglomerative Clustering, Gaussian Mixture Models (GMM), and K-means. Additionally, a novel scoring method is introduced that assesses the efficacy of the algorithm in real time by utilising response time and distance measurements. The proposed system surpasses conventional methods by achieving a novel distance score of 7.581 and 95% accuracy through k-fold cross-validation.