An improved regularization based Lagrangian asymmetric ν-twin support vector regression using pinball loss function.
Umesh GuptaDeepak GuptaPublished in: Appl. Intell. (2019)
Keyphrases
- loss function
- support vector regression
- support vector
- solution path
- regularization term
- reproducing kernel hilbert space
- risk minimization
- hinge loss
- support vector classification
- stochastic gradient descent
- convex loss functions
- hybrid model
- pairwise
- kernel function
- logistic regression
- regression model
- optimal solution
- empirical risk
- support vector machine
- support vector machine svm
- image processing
- uniform convergence
- classification accuracy
- learning problems
- kernel methods
- kernel learning
- regularization parameter
- feature vectors
- hyperparameters
- generalization ability