Which technique is most effective for handling missing values in a small dataset with non-random missingness?
Key points
- Addresses non-random missingness
- Preserves feature relationships
- Reduces bias compared to simple imputation
Ready to go further?
Key points
Ready to go further?