Various Regression models including linear, polynomial, ridge, lasso and elastic net were experimented with to find which model best predicted health insurance costs. The models were evaluated using cross-validation, from which the best models were optimized using randomized search. The best model was then evaluated on the test data.