Perturbed Iterate SGD for Lipschitz Continuous Loss Functions.
Michael R. MetelAkiko TakedaPublished in: J. Optim. Theory Appl. (2022)
Keyphrases
- loss function
- stochastic gradient descent
- pairwise
- learning to rank
- squared error
- support vector
- logistic regression
- learning models
- loss minimization
- risk minimization
- original data
- hinge loss
- boosting algorithms
- convex loss functions
- update rules
- learning algorithm
- weight vector
- reproducing kernel hilbert space
- machine learning algorithms
- markov random field
- support vector machine