Trimmed Robust Loss Function for Training Deep Neural Networks with Label Noise.
Andrzej RusieckiPublished in: ICAISC (1) (2019)
Keyphrases
- loss function
- label noise
- neural network
- training process
- stochastic gradient descent
- pairwise
- active learning
- training algorithm
- support vector
- training set
- training data
- noise tolerant
- training examples
- data sets
- image processing
- learning theory
- markov random field
- special case
- noise model
- boosting algorithms
- objective function
- image segmentation
- risk minimization
- convex loss functions