Certifying Incremental Quadratic Constraints for Neural Networks via Convex Optimization.
Navid HashemiJustin RuthsMahyar FazlyabPublished in: L4DC (2021)
Keyphrases
- convex optimization
- neural network
- quadratic program
- convex constraints
- low rank
- interior point methods
- semi infinite
- total variation
- augmented lagrangian
- primal dual
- convex relaxation
- convex formulation
- constrained optimization
- convex optimization problems
- convex sets
- pairwise
- high quality
- alternating direction method of multipliers
- operator splitting
- linear constraints
- low rank matrix
- semi definite programming
- linear programming problems
- learning problems
- linear programming
- computational complexity
- genetic algorithm