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.
- Type of
- machine learning
- Instance of
- field of study, field of study
Connected
More concepts & methods
- extreme learning machinetype of artificial neural network
- automated machine learningprocess of automating the end-to-end process of machine learning
- structured predictionsupervised machine learning techniques
- lazy learningtype of machine learning method
- quantum neural networkcomputational neural network model based on the principles of quantum mechanics
- echo state networkrecurrent neural network with a sparsely connected hidden layer
Sources
- Wikidata Q96416030 (structured facts, CC0)
- Wikipedia