Short Answer

When does the random 70/30 split practice fail, according to the source?

Question: In one to three sentences, state the specific condition under which the source says a random 70%/30% train/test split is a bad idea.

Sample answer: The random 70/30 split is a bad idea when the training distribution differs from the distribution the system ultimately needs to perform well on, such as when training data is website images but the target use is mobile phone images.

Key points:

  • Failure condition is distribution mismatch
  • Training distribution vs. target-use distribution
  • Example: website images vs. mobile phone images

Rubric: Full credit requires stating the distribution-difference condition; partial credit for mentioning the split without identifying the mismatch condition.

0

1

Updated 2026-07-10

Contributors are:

Who are from:

Tags

Machine Learning

Deep Learning

Machine Learning Strategy

Supervised Learning

Dive into Deep Learning @ D2L

Data Science

Machine Learning Yearning @ DeepLearning.AI