Smart Emergency Alerting System: A Machine Learning Approach

Recently, Saudi Arabia has hosted significant sports and technology events. Saudi Arabia has also successfully secured the bid to host Expo 2030 and declared its intention to host the FIFA World Cup in 2034. These crowds pertain to the elderly and special needs groups, which constitute the most vulnerable category in general, and they usually require some form of assistance. This study proposes a framework that integrates beacon technology with deep learning and real-time camera feed analysis to help crowds with special needs request emergency assistance by training a prioritization model for a patient based on this labeled dataset of 6962 records that consists of demographic and clinical details. In this, the random forest gives a nearly perfect classification and is the most successful model compared to the logistic regression and the SVM models. Logistic regression and the SVM model were not good with the minority classes. Once again, the model has been used by the application of that integrated application, and the result will be better in giving more priority to the medical emergency request and making a sound response. Another disadvantage could be, for example, accuracy in GPS coordinates, illumination conditions, and crowd density in indoor situations.