Asymptotic minimization of sequences of loss functions constrained by families of quasi-nonexpansive mappings and its application to online learning
Konstantinos SlavakisIsao YamadaPublished in: CoRR (2010)
Keyphrases
- loss function
- online learning
- pairwise
- support vector
- loss minimization
- squared error
- learning to rank
- logistic regression
- e learning
- empirical risk
- risk minimization
- objective function
- hinge loss
- pairwise constraints
- reproducing kernel hilbert space
- active learning
- convex loss functions
- update rules
- machine learning
- data points
- square loss