The Local True Weight Decay Recursive Least Square Algorithm

Authors:
Chi Sing Leung;Kwok-Wo Wong;Yong Xu
Affiliations:
Department of Electronic Engineering, City University of Hong Kong, Hong Kong,;Department of Electronic Engineering, City University of Hong Kong, Hong Kong,;Department of Electronic Engineering, City University of Hong Kong, Hong Kong,
Venue:
Neural Information Processing
Year:
2007

Citing 8
Cited 0

Adaptive filter theory (2nd ed.)

Adaptive filter theory (2nd ed.)
Bayesian interpolation

Neural Computation
A practical Bayesian framework for backpropagation networks

Neural Computation
Original Contribution: Optimal filtering algorithms for fast learning in feedforward neural networks

Neural Networks
Optimal, predictive, and adaptive control

Optimal, predictive, and adaptive control
A pruning method for the recursive least squared algorithm

Neural Networks
On the Kalman filtering method in neural network training and pruning

IEEE Transactions on Neural Networks
Two regularizers for recursive least squared algorithms in feedforward multilayered neural networks

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

The true weight decay recursive least square (TWDRLS) algorithm is an efficient fast online training algorithm for feedforward neural networks. However, its computational and space complexities are very large. This paper first presents a set of more compact TWDRLS equations. Afterwards, we propose a local version of TWDRLS to reduce the computational and space complexities. The effectiveness of this local version is demonstrated by simulations. Our analysis shows that the computational and space complexities of the local TWDRLS are much smaller than those of the global TWDRLS.