Convergence conditions, line search algorithms and trust region implementations for the Polak-Ribière conjugate gradient method.
Luigi GrippoStefano LucidiPublished in: Optim. Methods Softw. (2005)
Keyphrases
- trust region
- conjugate gradient method
- line search
- conjugate gradient
- search algorithm
- convergence property
- convergence rate
- levenberg marquardt
- hessian matrix
- objective function
- faster convergence
- optimization methods
- training algorithm
- global optimum
- global convergence
- sufficient conditions
- newton method
- maximum likelihood estimation
- convergence analysis
- back propagation
- positive definite
- mean shift
- learning rate
- optimization method