MSN-03 // DEPLOYED
HEART DISEASE PREDICTION
~82% accuracy from five compared models, deployed as a Flask web app.
~82% Accuracy5 ModelsTop-3 EnsembleFlask App
BRIEFING
An end-to-end capstone comparing five classifiers (KNN, Decision Tree, Random Forest, SVM, and Logistic Regression), each also tested with bagging and AdaBoost, on the 13-feature clinical heart dataset. Random Forest was the strongest at ~82% accuracy; the three strongest models combine by majority vote for a ~82% ensemble. The model was wrapped in a Flask web app for real-time risk scoring — moving it from notebook to something a user can actually interact with.
ROLE / METHOD / OUTCOME
Role
Capstone lead.
Method
Five-model comparison with a top-3 majority-vote ensemble, served via Flask.
Outcome
~82% Random Forest accuracy; ensemble just behind.
STACK: Python · scikit-learn · Flask