Project dossier
Multiple Disease Prediction
Machine Learning / Health
- PythonCore language
- Scikit-learnClassification models
- StreamlitMulti-page web UI
- Pandas + NumPyPreprocessing
- PickleModel serialization for runtime inference
01 // What it is
A machine-learning web app that predicts the likelihood of multiple diseases — diabetes, heart disease, and Parkinson's — from user-entered health metrics, built as a multi-page Streamlit interface over pre-trained models.
02 // Why it exists
An academic project used to practice the full supervised-learning loop end to end: source real datasets, preprocess, train and compare classifiers, serialize the winners, and put them behind an interface a non-technical person can actually use.
03 // Surfaces
Predictor
Diabetes
Predictor
Heart Disease
Predictor
Parkinson's
04 // Stack
- PythonCore language
- Scikit-learnClassification models
- StreamlitMulti-page web UI
- Pandas + NumPyPreprocessing
- PickleModel serialization for runtime inference
05 // Honest status
Honest status
Complete academic project. Models are trained offline on public Kaggle datasets, serialized, and loaded at runtime for inference. It is a learning exercise, not a medical device, and makes no clinical claims.