QuestionQ100

VerifiedML – Data

An engine-manufacturing facility wants to use machine learning to identify faulty bolts. Which of the following would introduce bias into the model?

Explanation

Deliberately omitting specific faulty conditions creates an unrepresentative training dataset. A model trained on that dataset can learn patterns that fail systematically for the excluded conditions, thereby reproducing the training-data bias in its predictions. AWS describes bias as an imbalance in training data and notes that insufficient representation in training data can be learned and reproduced by a model.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!