Diabetes Prediction Web App

Overview

This project is a machine learning-based web application that predicts whether a person is diabetic or not using medical input data. It uses a trained classification model served via a FastAPI backend.

Tech Stack Used

How It Works

1. Model Training (Offline):
   • Dataset used: PIMA Indian Diabetes Dataset.
   • Preprocessed data and trained a classifier using Scikit-learn.
   • Model saved using pickle.
2. API Integration (FastAPI):
   • Created a POST endpoint /diabetes_prediction to accept medical data.
   • The input is parsed using Pydantic's BaseModel.
   • Data is converted into a format that the model understands.
   • The trained model returns a prediction (diabetic or not diabetic).
3. Frontend:
   • A simple HTML form collects user input.
   • Sends the data to the FastAPI backend using fetch or axios.
   • Displays prediction results to the user.

Input Parameters

• Pregnancies
• Glucose
• Blood Pressure
• Skin Thickness
• Insulin
• BMI
• Diabetes Pedigree Function
• Age

Prediction Output

• the person is diabetic
• the person is not diabetic

Future Enhancements

• Add user authentication & login.
• Store prediction history in a database.