Machine Learning Based Electronic Triage for Emergency Department

Increase in the number of casualties visit to Emergency Department (ED) have lead to over crowd and delay in medical care. Hence, electronic triaging has been deployed to alleviate these problems and improve managing the patient. In this paper research methodology framework based on diagnostic and cross-sectional study is used for patient triage. The empirical approach is used to build models for patient triage to correctly predict the patient’s medical condition, given their signs and symptoms. Models are built with supervised learning algorithms. The “Naive Bayes”, “Support Vector Machine”, “Decision Tree” and, “Neural Network” classification models are implemented and evaluated using chi-square statistical test. This study infers the significance of using machine learning algorithms to predict patient’s medical condition. Support Vector Machine and Decision Tree have shown better performance for the considered dataset.