Improved generalization of neural classifiers with enforced internal representation

Authors:
Iveta Mrázová;Dianhui Wang
Affiliations:
Charles University, Faculty of Mathematics and Physics, Department of Software Engineering, Malostranské nám. 25, 118 00 Praha 1, Czech Republic;Department of Computer Science and Computer Engineering, La Trobe University, Melbourne, Vic. 3086, Australia
Venue:
Neurocomputing
Year:
2007

Citing 20
Cited 3

What size net gives valid generalization?

Neural Computation
Creating artificial neural networks that generalize

Neural Networks
Hints and the VC dimension

Neural Computation
Learning from hints

Journal of Complexity
The nature of statistical learning theory

The nature of statistical learning theory
Structural learning with forgetting

Neural Networks
Neural networks for pattern recognition

Neural networks for pattern recognition
Merging back-propagation and Hebbian learning rules for robust classifications

Neural Networks
Neurosmithing: improving neural network learning

The handbook of brain theory and neural networks
No free lunch for early stopping

Neural Computation
Rule extraction by successive regularization

Neural Networks
Information Retrieval

Information Retrieval
Machine Learning

Machine Learning
Mining the Web: Discovering Knowledge from HyperText Data

Mining the Web: Discovering Knowledge from HyperText Data
Evolving efficient learning algorithms for binary mappings

Neural Networks - 2003 Special issue: Advances in neural networks research — IJCNN'03
Data smoothing regularization, multi-sets-learning, and problem solving strategies

Neural Networks - 2003 Special issue: Advances in neural networks research — IJCNN'03
Letter: Learning similarity for semantic images classification

Neurocomputing
The “weight smoothing” regularization of MLP for Jacobian stabilization

IEEE Transactions on Neural Networks
Improving generalization performance using double backpropagation

IEEE Transactions on Neural Networks
Curvature-driven smoothing: a learning algorithm for feedforward networks

IEEE Transactions on Neural Networks

Dealing with limited data in ballistic impact scenarios: an empirical comparison of different neural network approaches

Applied Intelligence
An optimization methodology for machine learning strategies and regression problems in ballistic impact scenarios

Applied Intelligence
Extreme learning machine based genetic algorithm and its application in power system economic dispatch

Neurocomputing

Quantified Score

Hi-index	0.02

Visualization

Abstract

In standard BP-networks, hidden neuron outputs are usually spread over the whole interval (0,1). In this paper, we propose an efficient framework to enforce a transparent internal knowledge representation in BP-networks during training. We want the formed internal representations to differ as much as possible for different outputs. At the same time, the hidden neuron outputs will be forced to group around three possible values, namely 1, 0 and 0.5. We will call such an internal representation unambiguous and condensed. The performance of BP-networks with enforced internal representations will be examined in a case study devoted to semantic image classification.