Cross-validation and data leakage
Why honest validation matters, how data leakage inflates metrics, and which safeguards Chrometrica has and does not have.
Replicates and Group K-Fold
Why replicates of one sample must stay in one fold, and when to average them instead.
Interpreting permutation p-values
What the permutation test p-value means and what it does not.
Hard vs soft PLS-DA
Difference between hard and soft PLS-DA classification rules, and how alpha and gamma work in Chrometrica.