Bias-Variance Tradeoff

The fundamental tradeoff in statistical modeling where reducing a model's bias (systematic error from simplifying assumptions) increases its variance (sensitivity to training data fluctuations), and vice versa, with the optimal balance minimizing total prediction error.