A Riemannian smoothing steepest descent method for non-Lipschitz optimization on submanifolds.
Chao ZhangXiaojun ChenShiqian MaPublished in: CoRR (2021)
Keyphrases
- steepest descent method
- step size
- convergence speed
- optimization algorithm
- optimization problems
- evolutionary algorithm
- pointwise
- low dimensional
- shape analysis
- back propagation
- euclidean space
- convergence rate
- global optimization
- fitness function
- neural network
- computational complexity
- optimal solution
- multiscale
- similarity measure