Exercise: Undersampling

Practice applying the NearMiss undersampling technique to balance severely imbalanced training data in a binary classification context. Learn to configure parameters to achieve a specific class ratio, undersample the majority class, and verify data balance for improved entity resolution modeling.