Learning project · API
Iris classification API
A small end-to-end model comparison and deployment exercise.
The problem
Compare classifiers and make predictions available through an API.
My contribution
Compared Logistic Regression, SVM, KNN, and Random Forest, then deployed the selected model.
Engineering decisions
- Use cross-validation, GridSearchCV, PCA, and noise-injection checks.
Results & evidence
- Documented 96–97% cross-validation accuracy on Iris; included an API, web frontend, and tests.