Tree-loss function for training neural networks on weakly-labelled datasets.
Sergey DemyanovRajib ChakravortyZongyuan GeSeyedBehzad BozorgtabarMichelle PabloAdrian BowlingRahil GarnaviPublished in: ISBI (2017)
Keyphrases
- loss function
- neural network
- stochastic gradient descent
- training process
- pairwise
- training algorithm
- support vector
- logistic regression
- linear svm
- learning to rank
- empirical risk
- training dataset
- risk minimization
- back propagation
- benchmark datasets
- training set
- training examples
- regularization term
- reproducing kernel hilbert space
- training data
- markov random field
- support vector machine
- special case