High Accuracy Automated Differential Diagnosis using Machine Learning

This study proposes a method to automate differential diagnosis based on the symptoms provided by the patient. The aim of this work is to make it accessible to all so that even the medical practitioners can access this making their job of classifying the diagnosis easier. To achieve this, the proposed machine learning algorithm has been integrated with the designed website which can be accessed by the use from anywhere around the world. A support vector machine learning algorithm is implemented which drives the developed system in training, testing, and prediction of the diagnosis. The system provides attributes depicting the results in form of a heatmap which is based on the confusion matrix which is formed with the tests data trained and predicted. The performance of the proposed diagnostic system has been verified by using the data set with 100 entries and has given 95% accuracy.