Convex vs nonconvex approaches for sparse estimation: Lasso, Multiple Kernel Learning and Hyperparameter Lasso.
Aleksandr Y. AravkinJames V. BurkeAlessandro ChiusoGianluigi PillonettoPublished in: CDC/ECC (2011)
Keyphrases
- group lasso
- multiple kernel learning
- convex optimization
- elastic net
- regularized regression
- regression model
- efficient optimization
- variable selection
- convex relaxation
- low rank
- multi task
- multiple kernel
- multi label learning
- kernel methods
- model selection
- feature selection
- sparse representation
- linear combination
- contingency tables
- learning problems
- kernel learning
- support vector
- objective function
- learning models
- regularization term
- convex sets
- small number