Short Answer

What is the name for a model that is too simple to fit the patterns in the data well?

Question: A classifier has 18% training error and only slightly worse dev error, so the gap between the two is small. What term is used for this kind of behavior when the model does not fit the data well enough?

Sample answer: This is called underfitting.

Key points:

  • The model has high bias and low variance.
  • The usual term for this situation is underfitting.

Rubric: The student must identify the term 'underfitting' from the description of high bias and low variance.

0

1

Updated 2026-08-12

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