Short Answer

When to Add a Mislabeled Column

Question: When should you add a "Mislabeled" column to an error analysis spreadsheet, and what does it help you measure?

Sample answer: Add the column when you suspect that label errors may account for a meaningful share of the mistakes. The column helps you record the fraction of examples whose labels appear to be wrong.

Key points:

  • Add it when you think mislabeled data may be a significant issue.
  • Use it to track the fraction of examples with incorrect labels.

Rubric: To receive full credit, the answer must state that the column is added when mislabeled examples may be common, and that it is used to track the share of examples with incorrect labels.

0

1

Updated 2026-08-12

Contributors are:

Who are from:

Tags

Machine Learning

Deep Learning

Supervised Learning

Dive into Deep Learning @ D2L

Data Science

Machine Learning Strategy

Machine Learning Yearning @ DeepLearning.AI