A machine learning–powered web application that predicts the likelihood of diabetes from health parameters. Trained on the Pima Indians Diabetes Dataset and built for early, accessible detection.
Project Overview
This ML web application uses a trained scikit-learn model to predict whether a person is likely to have diabetes based on 8 health input parameters. The model was trained on the Pima Indians Diabetes Dataset, a well-known medical dataset in the ML community.
The user enters their health details, the model processes the inputs, and the app returns a prediction of Positive or Negative along with a confidence level. The goal is accessible early detection for people who may not have immediate access to comprehensive medical testing.
This project was also my deep-dive into the ML pipeline, data preprocessing, feature engineering, model selection, hyperparameter tuning, and Flask deployment.
Input Parameters
The model analyses 8 medical parameters to make its prediction.
ML Pipeline
Tech Stack