To begin learning about meta networks, we first need to define the following terms:
- Support set: Sampled input data points (x,y) from the training set.
- Test set: Sampled data points (x,y) from the training set.
- Embedding function (
): As part of a meta-learner, the embedding function is very similar to Siamese networks. It is trained to predict whether two inputs are of the same class. - Base-learner model (
): A base-learner model attempts to complete the actual learning task (for example, a classification model). - : Fast ...
