Learn Before
Case Study

Diagnose what a training-to-dev performance decline represents.

Case context: A team observes that its algorithm does better on the training set and worse on the dev set. It wants to interpret the difference using the supplied informal definition.

Question: What should the team compare, and what does the amount of worsening represent?

Sample answer: The team should compare the algorithm's dev-set performance with its training-set performance. The amount by which it does worse on the dev set than on the training set is informally regarded as the algorithm's variance.

Key points:

  • Compare dev-set performance with training-set performance.
  • Measure how much worse the dev result is.
  • Interpret that amount informally as variance.

Rubric: The response should name training and dev performance, preserve the direction of the comparison, and identify the resulting amount of worsening as informal variance.

0

1

Updated 2026-07-20

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