Skip to content

FRM Part I · FRM Exam Part I · Machine Learning and Prediction

A risk team has transaction data for 50,000 corporate clients with no labels. They run k-means to group clients with similar trading behavior into five clusters. Which statement about this approach is most accurate?

K-means clustering is unsupervised learning because it groups observations using only their features, with no target labels. Specifying the number of clusters is a hyperparameter choice, and iterating on cluster centers or naming clusters afterward does not turn it into supervised or reinforcement learning.

  1. AIt is supervised because the number of clusters is specified in advance
  2. BIt is unsupervised because the algorithm finds structure without target labelsCorrect
  3. CIt is reinforcement learning because cluster centers are updated iteratively
  4. DIt is supervised because the clusters can later be given names

Explanation

K-means uses only the features and no target variable, so it is unsupervised. Choosing k is a hyperparameter, not a label. Iterative updating and naming clusters afterward do not make it supervised or reinforcement learning.

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.

Practise Machine Learning and PredictionFree diagnostic mock

More Machine Learning and Prediction questions