Too Simple or Too Complex? The Bias–Variance Tradeoff in Machine Learning

The central question in supervised machine learning is not whether a model can fit the data it has...