Skip to content

FRM Part I · FRM Exam Part I · Machine-Learning Methods

An analyst tunes a regularization penalty in a ridge-type regression. As the penalty parameter is increased from a very small value to a very large value, which of the following describes the expected change in the model's bias and variance?

Bias increases and variance decreases. A larger regularization penalty shrinks coefficients and constrains model flexibility, so predictions become more stable across samples but are systematically further from the true relationship. This is the standard bias-variance tradeoff effect of stronger regularization.

  1. ABias decreases and variance increases
  2. BBias increases and variance decreasesCorrect
  3. CBoth bias and variance decrease
  4. DBoth bias and variance increase

Explanation

A larger penalty shrinks coefficients toward zero, making the model less flexible. Less flexibility means the fitted function is more stable across samples (lower variance) but deviates more systematically from the true relationship (higher bias).

Did you get it right without looking?

One question tells you little. A timed set on Machine-Learning Methods shows your real accuracy, how long you take and where you lose marks.

Practise Machine-Learning MethodsFree diagnostic mock

More Machine-Learning Methods questions