Inspect real data, handle missing values, encode features, avoid leakage, and build repeatable transformations.
For a small example, five house records include one impossible negative area. Inspect ranges and missingness, identify whether the value is an entry error, and document rather than silently delete it.
Prepare mixed numeric and categorical customer data for churn prediction. Split first, fit imputers and encoders on training folds, transform validation data, and preserve feature lineage.
Initializing Interactive Playground...
Before running the pipeline, predict how the imputation value changes when held-out measurements are included and name why even a small numerical change violates the evaluation boundary.
The starter calculates one imputation value from training and held-out measurements together, so evaluation data changes the model pipeline before evaluation begins.
Check: The checks pass and your explanation names the leaked parameter, the protected boundary, and the production artifact that must be versioned.
Failing starter lab
Compute a training mean and use it to fill missing values in both train and test partitions without reading test values into the statistic.
Run the incomplete starter before opening hints. Use the first wrong output or failed check to form a repair hypothesis.
Explanations after each answer · 80% to pass · Data Exploration & Preprocessing
Chapter 1 of 4 · Core
Write a data dictionary that states what one row represents, when each feature becomes available, the unit and allowed range of every field, and how the label was produced. The same column name can mean different things at collection time and prediction time. This temporal distinction is central to leakage.
Separate identifiers from candidate features. An identifier may encode location, time, or acquisition process and let a model memorize entities. It can still be essential for grouped splitting and auditing.
Ideas to understand
Two free references selected for Data Exploration & Preprocessing.
Primary sources and free technical reading for Data Exploration & Preprocessing.
Official NumPy guide to arrays, shapes, axes, indexing, and vectorized operations.
Free reference for preprocessing, model selection, metrics, and classical ML algorithms.
Research paper proposing documentation of dataset motivation, composition, collection, and recommended uses.