Skip to content

CS Professional · Artificial Intelligence, Data Analytics and Cyber Security - Laws and Practice · Internet and Other Technologies

A Mumbai lending fintech trains a model on historical loan records where each past applicant is tagged as 'defaulted' or 'repaid'. The model then predicts the status of new applicants. Which machine learning approach is this?

This is supervised learning. The model is trained on historical records that already carry labels, defaulted or repaid, and learns to predict the label for new applicants. Unsupervised learning uses unlabelled data, and reinforcement learning relies on rewards from an environment.

  1. AUnsupervised learning, because it finds groups without labels
  2. BSupervised learning, because it learns from labelled outcomesCorrect
  3. CReinforcement learning, because it receives rewards from an environment
  4. DTransfer learning, because it is copied from a model built abroad

Explanation

The training data carries known labels (defaulted or repaid) and the model learns the mapping from applicant features to those labels, which is supervised learning. Unsupervised learning would work without labels, for example clustering customers. Nothing in the scenario involves rewards or reuse of a prior model.

Did you get it right without looking?

One question tells you little. A timed set on Internet and Other Technologies shows your real accuracy, how long you take and where you lose marks.

Practise Internet and Other TechnologiesFree diagnostic mock

More Internet and Other Technologies questions