← All explorers

Bias–Variance Tradeoff

Fit the same model to many random samples. Watch the fits scatter (variance) and drift from truth (bias) as you change complexity, data, and noise.

What changed

The shaded band is the spread of fits across many random samples — its height is variance. The solid line is their average; its gap from the dashed true curve is bias. Raise the degree, add data, or change noise and watch the two trade off.

Loading…

Illustrative model for building intuition, not a live system.

Read the theory: ml-fundamentals