Essay

Analyzing the Impact of Adding Training Data on Bias and Variance

Question: According to Andrew Ng's Machine Learning Yearning, under what conditions is adding training data viable to address high variance, and how does this action typically affect the model's bias?

Sample answer: Adding more training data is a viable solution to high variance as long as you have access to significantly more data and the computational power required to process it. This remedy is highly effective because adding training data usually reduces variance without affecting bias.

Key points:

  • Adding training data requires access to significantly more data.
  • Adding training data requires enough computational power to process the data.
  • Adding training data typically reduces variance.
  • Adding training data does not affect bias.

Rubric: The response must identify that adding data requires access to significantly more data and sufficient computational power. It must also explain that adding training data typically reduces variance while leaving bias unaffected.

0

1

Updated 2026-05-26

Contributors are:

Who are from:

Tags

Machine Learning

Deep Learning

Supervised Learning

Dive into Deep Learning @ D2L

Data Science

Machine Learning Strategy