Short Answer

What training data change helps with a data mismatch problem?

Question: A classifier is doing poorly on a development set because the development examples come from a different situation than the training data. What is one practical change to the training data that can help?

Sample answer: One practical response is to collect additional training examples that look more like the difficult development examples the system is failing on.

Key points:

  • Add more training data.
  • Choose data that resembles the hard development cases.

Rubric: The answer must say to obtain more training examples that match the challenging development examples.

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