Partial BFGS Update and Efficient Step-Length Calculation for Three-Layer Neural Networks

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Deep learning in neural networks: An overview (2014, methodology): “Least-squares methods (Gauss– Newton, Levenberg–Marquardt) (Gauss, 1809; Levenberg, 1944; Marquardt, 1963; Newton, 1687; Schaback & Werner, 1992) and quasi-Newtonmethods (Broyden–Fletcher–Goldfarb–Shanno, BFGS) (Broyden et al., 1965; Fletcher & Powell, 1963; Goldfarb, 1970; Shanno, 1970) are computa”
On supervised learning from sequential data with applications for speech regognition (1999, methodology): “Neural networks for supervised learning from sequences(2) second order methods (Shepherd, 1997), which also use the second derivative(Hessian) or approximations to it, for example quasi-Newton, conjugate gradient,Levenberg-Marquardt, BPQ (Saito and Nakano, 1997) or Quickprop (Fahlmann,1988).”
Normalized Mutual Information Feature Selection (2009, unknown): “To compute , a three-layered MLP classifier is trained to minimize the sum of square errors by using a secondorder backpropagation quasi-Netwon (BPQ) learning algorithm [39].”
Optimal Linear Combination of Neural Networks for Improving Classification Performance (2000, methodology): “We used a three-layered feedforward neural network with 15 hidden units and used the BPQ algorithm [ 18 ] to train the neural networks.”