Why are instance-based learning algorithms sometimes referred to as "lazy learning" algorithms?
A) Because they require minimal computational resources during training.
B) Because they defer the majority of the computation to the prediction phase, only calculating predictions when needed.
C) Because they rely heavily on pre-existing knowledge stored in the model.
D) Because they prioritize speed over accuracy in model training.
