When selecting base models for stacking, what is a crucial consideration?
High accuracy on the training set
Similar predictive performance
Diversity in model architecture and training data
Low computational complexity

Machine Learning Algorithms Exercises are loading ...