A neighborhood-based enhancement of the Gauss-Newton Bayesian regularization training method

Authors:
Miguel Pinzolas;Ana Toledo;Juan Luís Pedreño
Affiliations:
Departamento de Ingeniería de Sistemas y Automática, Universidad Politécnica de Cartagena, Cartagena, Spain;Departamento de Tecnología Electrónica, Universidad Politécnica de Cartagena, Cartagena, Spain;Departamento de Tecnologías de la Información y las Comunicaciones, Universidad Politécnica de Cartagena, Cartagena, Spain
Venue:
Neural Computation
Year:
2006

Citing 5
Cited 1

Introduction to non-linear optimization

Introduction to non-linear optimization
Neural networks and the bias/variance dilemma

Neural Computation
A practical Bayesian framework for backpropagation networks

Neural Computation
Neighborhood based Levenberg-Marquardt algorithm for neural network training

IEEE Transactions on Neural Networks
Improvement of the neighborhood based Levenberg-Marquardt algorithm by local adaptation of the learning coefficient

IEEE Transactions on Neural Networks

Neighborhood based modified backpropagation algorithm using adaptive learning parameters for training feedforward neural networks

Neurocomputing

Quantified Score

Hi-index	0.00

Visualization

Abstract

This work develops and tests a neighborhood-based approach to the Gauss-Newton Bayesian regularization training method for feedforward backpropagation networks. The proposed method improves the training efficiency, significantly reducing requirements on memory and computational time while maintaining the good generalization feature of the original algorithm. This version of the Gauss-Newton Bayesian regularization greatly expands the scope of application of the original method, as it allows training networks up to 100 times larger without losing performance.