The Empirical Duality Gap of Constrained Statistical Learning.
Luiz F. O. ChamonSantiago PaternainMiguel Calvo-FullanaAlejandro RibeiroPublished in: ICASSP (2020)
Keyphrases
- statistical learning
- duality gap
- statistical learning theory
- primal dual
- model selection
- augmented lagrangian
- machine learning
- supervised learning
- set covering problem
- information theory
- convex programming
- semi supervised learning
- text mining
- optimal solution
- inequality constraints
- saddle point
- optimality conditions
- dual formulation
- convex optimization
- active learning
- evolutionary algorithm
- lagrange multipliers
- image processing
- decision trees
- multiscale
- dimensionality reduction
- manifold learning
- image restoration