DISEASE RISK PREDICTION USING AUTOMATED DIAGNOSIS SYMPTOMS

Abstract

Health risk prediction in today's world plays an important role by identifying risks from various diseases. Predicting these diseases helps prevent future risks. We approached the dataset, which is a structured collection designed for training and evaluating Machine Learning (ML) models in the field of automated disease diagnosis. There are 41 diseases linked to 377 symptoms. Many methods exist for treating different diseases worldwide. In this paper, we designed a disease prediction system that takes the symptoms provided by an individual as input and produces the predicted output, which is the most likely disease. This system serves as a valuable resource for examining ML algorithms. We used Random Forest to determine the most accurate and effective predictive model, which achieved an accuracy of 83.85%. By building and testing these automated systems, we aim to create a helpful tool for doctors. This means they can identify potential diseases earlier and more accurately, ultimately reducing diagnostic time and providing better care to patients more quickly.