The problem Disease Prediction using ML solves

The webiste guides the user to two different platforms which focuses on predicting if their given health conditions shows signs of diabetes or heart disorders. The main webiste has two links (which can be extended subjected to more datasets).

The first option guides the user to preidct if the given health conditions are a sign of diabetes or not. The required medical details for finding diabetes are: Pregnancies, Glucose, Blood Pressure, Skin Thickness, Insulin, BMI, DPF, Age.

The second option guides the user to predict if the given health consitions are a sign of some kind of heart disorders or not. The required medical details for finding heart disorders are: Medical Data Required: Age, Sex, Chest Pain, Rest Blood Pressure, Cholestrol, Fasting Blood Sugar, Rest ECG, Maximum Heart Rate, Exercise Induced Angina, ST Depression induced by exercise relative to rest, Slop, No. of major vessels, Thal.

Challenges we ran into

The biggest challenge faced by the team was when we tried to deploy the webisite which shows the ML predictions. We experimented on various methodologies such as trying to host with github pages and trying with default steamlit.share options.

Finally, we decided to create 3 repos for this project, which lead to hosting 3 seperate websites. One link shows the prediction of heart diseases, one link shows the prediction of diabetes and the last link is the main website which is the home page of our team and this website is the one which leads the user to which ever prediction platform he wants to land on to.

For hosting the ML prediction sites, we finally decided to use heroku and for hosting the main home website, we used digital ocean.