A Theoretical and Empirical Comparison of Gradient Approximations in Derivative-Free Optimization.
Albert S. BerahasLiyuan CaoKrzysztof ChoromanskiKatya ScheinbergPublished in: Found. Comput. Math. (2022)
Keyphrases
- derivative free
- gradient method
- unconstrained optimization
- neural learning
- convergence rate
- optimization methods
- constrained optimization
- fan beam
- negative matrix factorization
- closed form
- search methods
- global optimization
- trust region
- incremental learning
- nonlinear optimization
- least squares
- genetic algorithm
- approximation methods
- edge detection