Feature Engineering and Preprocessing

Apply feature scaling, normalization, and log transformations to prepare data for modeling.

Apply appropriate imputation strategies for missing data, explaining trade-offs.

Encode categorical variables using one-hot, label, and target encoding, articulating the pros and cons of each.

Detect and treat outliers using statistical methods (e.g., IQR, Z-score) and domain knowledge.

Engineer features from temporal data, such as lags and rolling windows.

Engineer features from transactional data, such as RFM metrics and behavioral aggregations.