Skip to content

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

Brevik Logistics introduced a demand forecasting model that performed well for two years. After a major shift in customer buying behaviour following a market disruption, forecast accuracy fell sharply, although the code has not changed. Which issue best explains this?

Model drift explains it. The model was trained on past buying patterns, and when customer behaviour changed the relationships it learned no longer held, so accuracy fell although the code was unchanged. Ongoing monitoring and retraining are needed to manage this.

  1. AModel drift, because the relationships in current data differ from those in the training dataCorrect
  2. BInsufficient processing power in the data centre
  3. COver-reliance on unstructured data
  4. DLack of a data protection policy

Explanation

Models learn patterns from historical data. When the real-world relationships change, performance degrades without any change in code. This is model (concept) drift, and it requires monitoring and retraining. Processing power would affect speed, not accuracy after a behavioural shift.

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