Case Study

Choosing the Easier Image Classification Task

Case context: You are building a wildlife photo app with a small budget for manual labeling. You must decide whether one component should learn to classify whether an image contains a vehicle or whether it contains a scooter.

Question: Based on the idea that simpler tasks need less data, which subtask should you choose for this component, and why?

Sample answer: Choose the task of deciding whether an image contains a vehicle. That is a broader, easier binary classification problem than deciding whether it contains a scooter. Because simpler tasks generally require fewer labeled examples, it is the better fit when labeling capacity is limited.

Key points:

  • Pick the 'contains a vehicle' task.
  • State that it is easier than scooter-level identification.
  • Explain that easier tasks need fewer training examples, which matches the data constraint.

Rubric: The response must select the easier task ('contains a vehicle') and justify it by stating that easier tasks require fewer training examples, making it appropriate under a limited labeling budget.

0

1

Updated 2026-08-12

Contributors are:

Who are from:

Tags

Python Programming Language

Data Science

Machine Learning

Deep Learning

Supervised Learning

Dive into Deep Learning @ D2L

Machine Learning Strategy

Machine Learning Yearning @ DeepLearning.AI