Skip to content

CS Professional · Artificial Intelligence, Data Analytics and Cyber Security - Laws and Practice · Artificial Intelligence - Introduction and Basics

A Bengaluru retailer's AI team splits its customer data into a training set, a validation set and a test set. What is the main purpose of the separate test set?

The test set gives a final, unbiased estimate of how the model performs on unseen data. It is kept aside until development ends. Tuning hyperparameters is done on the validation set, and using the test set for that would leak information and overstate performance.

  1. ATo tune the model's hyperparameters repeatedly during development
  2. BTo give a final unbiased estimate of performance on data the model has never seenCorrect
  3. CTo enlarge the training data when the model underfits
  4. DTo clean missing values before training begins

Explanation

The test set is held back until the end so that it measures generalisation on unseen data. Hyperparameter tuning is the job of the validation set; using the test set for it would leak information and make the estimate optimistic.

Did you get it right without looking?

One question tells you little. A timed set on Artificial Intelligence - Introduction and Basics shows your real accuracy, how long you take and where you lose marks.

Practise Artificial Intelligence - Introduction and BasicsFree diagnostic mock

More Artificial Intelligence - Introduction and Basics questions