DataX

Practice session

A data scientist observes that a gradient boosting model achieves 98% accuracy on training data but only 72% on held-out test data. Which regularization approach MOST directly addresses this overfitting?

Loading your practice session…