Zero-Shot Transfer Learning Algorithm
(Redirected from zero-shot learning algorithm)
Jump to navigation
Jump to search
A Zero-Shot Transfer Learning Algorithm is a transfer learning algorithm that can be implemented by a zero-shot learning system to solve zero-shot learning tasks.
- Context:
- ...
- Example(s):
- Direct Attribute Prediction: An algorithm that learns to map input features to a semantic space, typically defined by class attributes.
- Indirect Attribute Prediction: An algorithm that first learns to map input features to the known class labels, and then maps class labels to the semantic space.
- Transductive Zero-Shot Learning Algorithm: An algorithm that leverages the knowledge of the test data (without class labels) during training.
- ...
- Counter-Example(s):
- See: Semantic Space, Auxiliary Information.