Zero-shot learning

problem setup in machine learning, where at test time, a learner observes samples from classes that were not observed during training, and needs to predict the class they belong to

Pronunciation

Categories