Skip to content

IAI Actuarial Core Principles · Risk Modelling and Survival Analysis · Elementary principles of machine learning

An analyst at an Indian insurer uses 5-fold cross-validation on 1,000 policies to choose a tuning parameter. How many policies are used to train the model in each fold, and how many times is each policy used for validation?

In 5-fold cross-validation of 1,000 policies, each fold has 200 policies. The model trains on the other four folds, so 800 policies, and each policy is in the validation fold exactly once across the five rounds.

  1. A800 for training; each policy validated onceCorrect
  2. B800 for training; each policy validated five times
  3. C200 for training; each policy validated once
  4. D1,000 for training; each policy validated once
  5. 750 for training; each policy validated four times

Explanation

With 5 folds each of 200 policies, every fold is held out once while the other four folds (800 policies) train the model. Hence each policy appears in a validation fold exactly once. Option with 200 reverses the roles of training and validation.

Did you get it right without looking?

One question tells you little. A timed set on Elementary principles of machine learning shows your real accuracy, how long you take and where you lose marks.

Practise Elementary principles of machine learningFree diagnostic mock

More Elementary principles of machine learning questions