Login / Signup
Powell's dogleg trust-region steps with the quasi-Newton augmented Hessian for neural nonlinear least-squares learning.
Eiji Mizutani
Published in:
IJCNN (1999)
Keyphrases
</>
trust region
nonlinear least squares
newton method
column generation
neural network
hessian matrix
learning algorithm
lower bound
optimization methods
feature space
back propagation
sensitivity analysis
regularized least squares