Skip to content

ACCA Strategic Professional · Strategic Business Leader · Machine learning, AI and robotics

Calder Bank wants to reduce card fraud losses. Fraud patterns change constantly and are not fully captured by any fixed list of rules. The bank holds ten years of labelled transactions marked genuine or fraudulent. Which approach best reflects how machine learning would add value?

A supervised machine learning model trained on the labelled transaction history is best. It learns patterns that distinguish fraud from genuine activity and can adapt to changing methods, unlike fixed rules, robotic copying of data or manual review, which cannot detect evolving fraud at scale.

  1. ATrain a supervised model on the labelled history so it learns patterns and flags new suspicious transactionsCorrect
  2. BHard-code a list of rules based on known frauds and review them annually
  3. CUse robotic process automation to copy flagged transactions into a spreadsheet
  4. DAsk staff to manually review every transaction above a set value

Explanation

Labelled historic data suits supervised learning, which can detect evolving patterns that fixed rules miss. Fixed rules cannot adapt between annual reviews. RPA only moves data and does not detect anything, and manual review is slow and not scalable.

Did you get it right without looking?

One question tells you little. A timed set on Machine learning, AI and robotics shows your real accuracy, how long you take and where you lose marks.

Practise Machine learning, AI and roboticsFree diagnostic mock

More Machine learning, AI and robotics questions