Learn Before
Short Answer

Informal Bias and Training Error

Question: In an informal explanation of bias, what training-set quantity is used as a quick proxy for bias, and when is that shortcut usually sensible?

Sample answer: A common informal proxy for bias is the model's error on the training set. This approximation is most reasonable when the training set is very large, so the training error is stable enough to reflect the model's fitting quality.

Key points:

  • Use training-set error as the informal proxy for bias.
  • The approximation is reasonable when the training set is very large.

Rubric: The answer should state that bias is informally approximated by training-set error and note that this works best with a very large training set.

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