Prepare data and features
Inspect examples, handle missing values, and transform inputs without leaking information.
- Explain features, preprocessing, and how to avoid data leakage.
Features are the input variables used to make a prediction. Inspect their types, ranges, missing values, and how they were collected. Transformations such as scaling, encoding categories, or filling missing values should be learned using training data only, then applied consistently to validation and test data. This prevents data leakage, where information unavailable at prediction time makes evaluation look better than it should.
A small example
1train_values = [10, 20, 30]
2train_mean = sum(train_values) / len(train_values)
3new_value = 40
4print("training mean:", train_mean)
5print("centered new value:", new_value - train_mean)training mean: 20.0 centered new value: 20.0
A feature must be available when the system makes its prediction. A column recorded after the outcome, such as a cancellation reason, can leak the answer into training. Keep preprocessing in a repeatable pipeline so training and production use the same transformations.
Key takeaways
Explain features, preprocessing, and how to avoid data leakage.
Evaluate on relevant unseen data and monitor the system after deployment.
Lesson quiz
5 questions · pass with 4 correct · up to 50 XP
Passing this quiz completes the lesson and keeps your streak going. Questions you miss come back in review sessions later.
Questions about this lesson
Stuck? Ask. Figured something out? Share it. Explaining is one of the best ways to learn.
Loading posts…