Paper Contents
Abstract
Road accidents continue to cause a high rate of injuries and deaths across the world, making them a pressing challenge for public safety. A practical solution to reduce the delay in emergency care is to strategically locate ambulances in advance, rather than dispatching them only after requests are made. This work presents a deep learningdriven clustering model designed to identify suitable positions for ambulance placement. Accident frequency is closely tied to regional factors and spatial patterns; therefore, maintaining these relationships during model development is essential for producing accurate real-time outcomes. To address this, the study employs a representation technique known as Cat2Vec, which enables the preservation of such patterns. The proposed framework is compared with widely used clustering methods, Including K-Means, Gaussian Mixture Models (GMM), and Agglomerative Clustering. In addition, a new scoring metric is introduced to measure algorithm performance with respect to both travel distance and response time. Experimental evaluation shows that the system achieves 95% accuracy using k-fold validation and records a distance score of 7.581, demonstrating clear improvements over conventional clustering techniques for ambulance deployment
Copyright
Copyright © 2025 Putti Tanuja . This is an open access article distributed under the Creative Commons Attribution License.
