Question: Explain why it is useful to search for clues in machine learning projects, and how building a simple baseline system helps.

Sample answer: Searching for clues is useful because they point to the changes that are likely to help and the changes that are unlikely to matter. That means teams can focus their time on high-value experiments instead of guessing, which can prevent long periods of wasted work. A simple baseline system helps because it gives a concrete reference point for studying errors. By examining where the baseline succeeds and fails, the team can identify patterns in the data, the labels, or the task itself, and those patterns often reveal the best next steps.

Key points:

• Clues help teams decide which ideas are promising
• Finding the right clues can save substantial development time
• A simple baseline exposes useful patterns in errors and successes
• Those patterns guide the next set of experiments

Rubric: The essay must explain that clues reduce wasted effort by pointing to promising directions and that a baseline system helps reveal those clues through its error patterns.