IAI Actuarial Core Principles · Risk Modelling and Survival Analysis · Elementary principles of machine learning
A binary classifier for insurance claim fraud is tested on 200 claims. It flags 40 as fraud, of which 30 are truly fraud. In total 50 of the 200 claims are truly fraud. What is the recall (sensitivity) for the fraud class?
Recall is 0.60. It is the proportion of actual fraud cases correctly identified: 30 true positives divided by 50 actual fraud claims. The value 0.75 is precision, which divides by the 40 flagged claims instead.
- A0.75
- B0.60Correct
- C0.20
- D0.85
- 0.15
Explanation
Recall = true positives / actual positives = 30/50 = 0.60. Precision would be 30/40 = 0.75, which is the first option and uses flagged claims as the base, the wrong denominator for recall.
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.
More Elementary principles of machine learning questions
- A general insurer in India wants to group its motor policyholders into segments using only rating variables such as vehicle age, engine capa…
- A general insurer wants to predict the claim amount (in rupees) for each motor policy from rating factors such as vehicle age, engine capaci…
- A lasso (L1-penalised) regression is fitted to predict lapse rates and the penalty parameter lambda is increased from a small value to a ver…
- A life insurer in Mumbai has data on 20,000 customers (age, premium paid, number of policies, channel) but no outcome variable. An analyst a…
- 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 pol…
- An actuary fits a flexible model to 200 motor claims and finds it has a very small error on those 200 records but a much larger error on 100…