An Empirical Study of Derivative-Free-Optimization Algorithms for Targeted Black-Box Attacks in Deep Neural Networks.
Giuseppe UghiVinayak AbrolJared TannerPublished in: CoRR (2020)
Keyphrases
- black box
- derivative free
- neural network
- optimization problems
- neural learning
- hybrid systems
- white box
- unconstrained optimization
- optimization methods
- black boxes
- test cases
- evolutionary algorithm
- combinatorial optimization
- constrained optimization
- artificial neural networks
- integration testing
- learning algorithm
- back propagation
- gradient method
- trust region
- optimization algorithm
- objective function