Short Answer

Why should training results be checked before expecting strong dev/test results?

Question: Answer in one to three sentences using the source's relationship between seen and unseen examples.

Sample answer: Training results show how the algorithm performs on examples it has seen, and dev/test performance is usually worse. Therefore, the algorithm must first perform well on the training set before strong results on unseen dev/test examples can be expected.

Key points:

  • Training examples are seen examples.
  • Dev/test performance is usually worse.
  • Good training performance comes first.

Rubric: The answer should mention seen training examples, the usual disadvantage of dev/test performance, and the need for good training performance first.

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