Groups
Bootstrap is a resampling method that estimates uncertainty by repeatedly sampling with replacement from the observed data.
The bias–variance tradeoff explains how prediction error splits into bias squared, variance, and irreducible noise for squared loss.