Question: Explain how having labeled input-output examples affects the kinds of outputs an end-to-end deep learning system can learn directly. Why is that important?

Sample answer: When a training set contains the right input-output pairs, an end-to-end model can learn to map inputs straight to structured, high-complexity outputs. That matters because the target does not have to be a single score or class label; it can also be a sequence, an image, or another rich data object. This expands the range of problems that machine learning can solve without breaking the task into separate hand-designed pieces.

Key points:

• The model needs matching labeled examples of inputs and desired outputs.
• It can learn outputs that are more complex than one numeric value.
• Examples of such outputs include text, images, and audio.

Rubric: A strong response should explain the role of labeled pairs and state that end-to-end systems can learn richer structured outputs than a single number, with at least one correct example such as text or images.