Category
Level
The bias–variance tradeoff explains how prediction error splits into bias squared, variance, and irreducible noise for squared loss.