Short Answer

Starting error analysis without predefined categories.

Question: According to Machine Learning Yearning, should you be concerned if you begin your error analysis without any predefined error categories in mind? Briefly explain the recommended initial step to generate categories.

Sample answer: No, you should not worry if you start with no categories. The recommended first step is to simply look at a couple of misclassified images, which will typically inspire a few ideas for initial error categories.

Key points:

  • Do not worry about lacking initial categories.
  • Look at a few images to generate category ideas.

Rubric: 1 point for stating not to worry about starting without categories; 1 point for explaining that looking at a few initial images will generate ideas for categories.

0

1

Updated 2026-05-27

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