SpaceX Falcon 9 Landing Prediction

The capstone for the IBM Data Science Professional Certificate. Launch data was collected from the SpaceX API and scraped from Wikipedia, cleaned and prepared with Pandas and NumPy, explored with SQL (SQLite) and visualizations, and launch sites were mapped with Folium. Four classifiers (logistic regression, SVM, decision tree and KNN) were tuned with cross-validated grid search. Validation accuracy ranged from about 0.85 to 0.89; test accuracy was 0.83 on a small 18-sample test set, so the comparison between models is indicative rather than conclusive.