Smooth Loss Functions for Deep Top-k Classification.
Leonard BerradaAndrew ZissermanM. Pawan KumarPublished in: ICLR (Poster) (2018)
Keyphrases
- loss function
- support vector
- hinge loss
- classification accuracy
- partially labeled data
- feature extraction
- feature selection
- risk minimization
- squared error
- logistic regression
- text classification
- machine learning
- loss minimization
- pairwise
- learning to rank
- boosting algorithms
- decision trees
- feature vectors
- support vector machine
- class labels
- bayes rule
- convex loss functions
- pattern classification
- high dimensional
- reproducing kernel hilbert space
- training set
- machine learning algorithms
- support vector machine svm