FRM Part I · FRM Exam Part I · Machine Learning and Prediction
A model predicts whether a trade is fraudulent. Of 1,000 trades, 20 are actually fraudulent. The model flags 25 trades, of which 15 are actually fraudulent. What are the model's precision and recall for fraud?
Precision is 60% and recall is 75%. Precision is true positives divided by flagged trades, 15 of 25, while recall is true positives divided by actual fraudulent trades, 15 of 20. The reversed pairing results from using the wrong denominators.
- APrecision 60%, recall 75%Correct
- BPrecision 75%, recall 60%
- CPrecision 60%, recall 15%
- DPrecision 75%, recall 75%
Explanation
True positives are 15. Precision = TP/flagged = 15/25 = 60%. Recall = TP/actual fraud = 15/20 = 75%. Swapping the denominators gives the 75%/60% distractor.
Did you get it right without looking?
One question tells you little. A timed set on Machine Learning and Prediction shows your real accuracy, how long you take and where you lose marks.
More Machine Learning and Prediction questions
- An analyst runs PCA on three unstandardized variables whose variances are 100, 4 and 1, with small covariances between them. The analyst not…
- A single-output neural network uses squared-error loss L = 0.5(y - ŷ)², where ŷ = w·x with no activation and no bias. For one observation, x…
- Two variables are standardized and have a correlation of 0.60. What are the eigenvalues of their correlation matrix, and what proportion of …
- A bank's credit-default classifier is tested on 200 loans. It flags 50 loans as defaults, of which 40 actually defaulted. In total 60 loans …
- A neural network with many parameters achieves very low error on the training set but substantially higher error on a validation set. Which …
- A risk manager clusters customers on two variables: annual transaction volume (in USD, ranging 0 to 2,000,000) and number of late payments (…