The problem Diabetes Predictor App solves

The application solves the biggest concerns of all PATHOLOGY CENTERS that how to infer the test results without getting the COSTLY appointments with doctors .
When a person gets his sugar test results he is completely OBLIVIOUS on how to infer whether he is DIABETIC OR NOT . from the data present in the TEST-Report which is full of MEDICAL-JARGONS.
The applicaion that I have created can solve this isssue by bridging the Knowlegde gap, using KAGGLE -DATASET TRAINED POLYNOMIAL REGRESSION MACHINE LEARNING MODEL, which is available for use live on internet!

Challenges I ran into

Getting accurate dataset was a challenge- Using 100% dataset present on Kaggle solved that for me.
Analysing data was difficult- Pandas, numpy, plotlib libraries of Pyhton helped me to analyse data and choose suitable model.
Making MLmodel- Using polynomial regresssion model of sklearn helped me to set up POLYNOMIAL REGRESSSION model
Training the model on test data set- Used sklearn library to divide test and train data in 20 / 80 ratio, for testing and training.
Deploying the model-LIVE! -- used Heroku server to host application by intgrating the model into FLASK(pyhton) Backend.