Case Study

Diagnose why adding internet cat images could help a mobile app's cat classifier.

Case context: A team building a mobile app to identify cat photos decides to add 20,000 internet cat images to their training set, in addition to their existing user-uploaded mobile-app images.

Question: Based on the source's explanation, diagnose why this addition of internet images is likely to help the app's classifier perform better on user-uploaded mobile-app images.

Sample answer: The additional internet images give the network more examples of what cats do and do not look like. Since internet images and user-uploaded mobile-app images share some similarities, the network can apply some of the knowledge it acquires from the internet images to the mobile-app images, improving its ability to classify them correctly.

Key points:

  • Additional internet images increase the number of cat examples the network sees
  • Internet images and mobile-app images share some similarities
  • This similarity allows knowledge acquired from internet images to transfer to mobile-app images

Rubric: Full credit for identifying that more examples improve the network's understanding of cats and that shared similarities between the two image sources allow transfer of that knowledge to mobile-app images.

0

1

Updated 2026-07-11

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