Question 1

What is the bias-variance tradeoff?

Accepted Answer

The bias-variance tradeoff is the balance between underfitting (high bias) and overfitting (high variance). As model complexity increases, bias decreases but variance increases, leading to a U-shaped test error curve.

Question 2

How is total error decomposed?

Accepted Answer

Total error = Bias² + Variance + Irreducible noise. Bias² measures average deviation from true function, variance measures sensitivity to training data, and noise is inherent randomness.

Question 3

What is the optimal complexity?

Accepted Answer

The optimal complexity minimizes test error. It is the point where adding more parameters increases variance more than it reduces bias.

⚡ Bias–Variance Tradeoff