The conjugate gradient algorithm on well-conditioned Wishart matrices is almost deteriministic.

Percy Deift Thomas Trogdon

Published in: CoRR (2019)

Keyphrases

conjugate gradient algorithm
positive definite
conjugate gradient
learning rate
evolutionary programming
regularization framework
levenberg marquardt
neural network
kernel function
artificial neural networks
reproducing kernel hilbert space
kernel machines
training algorithm
global optimization
search direction