Similar Symptoms Prediction and Disease Prediction Using Machine Learning Techniques

Abstract

Due to the increase in the amount of data growth in the medical and healthcare field, data mining and machine learning technologies can play an important role in the prediction of diseases based on the existing data. Because the average waiting time for patients to meet with their doctor is about twenty to thirty minutes due to many reasons like the doctor may be running late or the patient is taking longer than the time allotted to him, etc. This can be a cause of frustration to the patients as they will have to wait for a considerable amount of time to meet their doctor. This problem can be solved by the proposed model. It makes use of the existing data in an attempt to solve the mentioned problem by significantly reducing the waiting time of the user as it is an application that can be used from the comfort of the home. The application not only provides easy diagnosis to help guide the user but also predicts the symptoms the user might have. The application has two steps. In the first step, a skip-gram model is used. In this model, the dataset is made into the symptom-disease format from the existing disease-symptom format. The symptoms of the disease are predicted in the first step. In the second step, these symptoms are fed into a model that can then predict the disease the person is suffering from. The symptoms of the diseases are predicted using the Random Forest, Decision tree, and Naive Bayes algorithms.