This graduate-level sequence explores advanced resampling methods for model validation, moving beyond simple train-test splits. Students implement k-fold, LOOCV, and nested cross-validation from scratch, while addressing critical challenges like data leakage and non-i.i.d. temporal data.