On Last-Layer Algorithms for Classification: Decoupling Representation from Uncertainty Estimation.
Nicolas BrosseCarlos RiquelmeAlice MartinSylvain GellyEric MoulinesPublished in: CoRR (2020)
Keyphrases
- classification algorithm
- feature representation
- benchmark datasets
- machine learning algorithms
- learning algorithm
- data structure
- theoretical analysis
- terms of classification accuracy
- significant improvement
- classification accuracy
- density estimation
- computationally efficient
- data sets
- continuous domains
- support vector machine
- machine learning methods
- preprocessing
- computational complexity
- pattern recognition
- decision trees
- feature selection
- pattern classification
- model selection
- classification models
- induction algorithms
- estimation problems
- classification rules
- decision rules
- text classification
- worst case
- multi class
- probability distribution
- training set
- neural network