Isotonic Modeling with Non-Differentiable Loss Functions with Application to Lasso Regularization.
Amichai PainskySaharon RossetPublished in: IEEE Trans. Pattern Anal. Mach. Intell. (2016)
Keyphrases
- loss function
- convex loss functions
- risk minimization
- reproducing kernel hilbert space
- loss minimization
- pairwise
- regularization term
- support vector
- learning to rank
- squared error
- gradient boosting
- empirical risk
- logistic regression
- bregman divergences
- hinge loss
- sparse representation
- feature selection
- feature space