Heart disease is the leading cause of death in the U.S. These high death rates can be offset by early detection and treatment of the condition. My motivation was to create a web app that would allow the user to input their vital readings and medical measurements to give them an idea as to whether or not they are susceptible to heart disease.
The web app returns as output a prediction as to whether or not the user has heart disease. The user is asked to input 14 features such as max resting heart rate, and blood pressure.
The web app was built with a react frontend and serviced with a python backend api which implemented a Random forest classification model using SKLearn. The model was trained using the Cleveland heart disease dataset: https://www.kaggle.com/ronitf/heart-disease-uci and the parameters were tuned using grid search. I tested multiple classification models such as SVC, naive bayes and logistical regression. Random forest had the highest accuracy at 86% on the testing set.
Log in or sign up for Devpost to join the conversation.