Skip to content

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

A data scientist uses 5-fold cross-validation on a dataset of 1,000 observations to select a model. Fold validation mean squared errors are 4.0, 5.0, 6.0, 5.0 and 5.0. Which statement is correct?

The cross-validation MSE estimate is 5.0, the average of the five fold errors, and each observation is used for validation exactly once. Each model is trained on the other four folds, which is 800 observations, not 200.

  1. AThe cross-validation estimate of MSE is 5.0, and each observation is used for validation exactly onceCorrect
  2. BThe cross-validation estimate of MSE is 5.0, and each model is trained on 200 observations
  3. CThe cross-validation estimate of MSE is 6.0, the largest fold error, to be conservative
  4. DThe cross-validation estimate of MSE is 5.0, and each observation is used for validation five times

Explanation

The average of the fold errors is (4+5+6+5+5)/5 = 25/5 = 5.0. In k-fold cross-validation each observation sits in the validation fold exactly once and in the training set k-1 times. Each model trains on 800 observations, not 200.

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