Scaling Ensemble Distribution Distillation to Many Classes with Proxy Targets.
Max RyabininAndrey MalininMark J. F. GalesPublished in: NeurIPS (2021)
Keyphrases
- training set
- spatial distribution
- ensemble methods
- ensemble learning
- mixture of gaussian distributions
- uniformly distributed
- data distribution
- base classifiers
- neural network
- training data
- power law
- random forest
- probability distribution
- ensemble classifier
- multiple targets
- multiple classifiers
- class membership
- expected error
- feature selection