Skip to content

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

A lapse-prediction model is fitted to 1,000 policies. Of the 100 policies that actually lapsed, the model flags 70 as lapses. Of the 900 policies that did not lapse, the model flags 90 as lapses. Treating lapse as the positive class, what are the model's recall (sensitivity) and precision?

Recall is 0.70 and precision is 0.4375. Recall is true positives divided by actual lapses, 70/100. Precision is true positives divided by all flagged policies, 70/(70+90) = 70/160. The two measures use different denominators, so they differ.

  1. ARecall 0.70, precision 0.4375Correct
  2. BRecall 0.70, precision 0.70
  3. CRecall 0.4375, precision 0.70
  4. DRecall 0.90, precision 0.4375
  5. Recall 0.70, precision 0.10

Explanation

True positives are 70 and false positives are 90, so 160 policies are flagged. Recall is 70/100 = 0.70, the share of actual lapses found. Precision is 70/160 = 0.4375, the share of flags that are correct. Swapping the two measures gives option 2 in reverse.

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