On Loss Functions for Deep Neural Networks in Classification.
Katarzyna JanochaWojciech Marian CzarneckiPublished in: CoRR (2017)
Keyphrases
- loss function
- neural network
- support vector
- pattern recognition
- hinge loss
- pairwise
- decision trees
- classification accuracy
- loss minimization
- feature extraction
- logistic regression
- machine learning
- learning to rank
- risk minimization
- feature vectors
- update rules
- kernel classifiers
- text classification
- squared error
- training process
- partially labeled data
- model selection
- supervised learning
- cost sensitive
- pattern classification
- benchmark datasets
- support vector machine svm
- bayes rule
- training samples
- similarity measure