Effectively extracting rules from trained neural networks based on the new measurement method of the classification power of attributes

Authors:
Dexian Zhang;Yang Liu;Ziqiang Wang
Affiliations:
School of Information Science and Engineering, Henan University of Technology, Zheng Zhou, China;Computer College, Northwestern Polytecnical University, Xi'an, P.R.C(China);School of Information Science and Engineering, Henan University of Technology, Zheng Zhou, China
Venue:
FSKD'05 Proceedings of the Second international conference on Fuzzy Systems and Knowledge Discovery - Volume Part II
Year:
2005

Citing 6
Cited 0

Extracting Refined Rules from Knowledge-Based Neural Networks

Machine Learning
Accelerating neural network training using weight extrapolations

Neural Networks
Rule-extraction by backpropagation of polyhedra

Neural Networks
NeuroRule: A Connectionist Approach to Data Mining

VLDB '95 Proceedings of the 21th International Conference on Very Large Data Bases
Extracting symbolic rules from trained neural network ensembles

AI Communications - Special issue on Artificial intelligence advances in China
Data dimensionality reduction with application to simplifying RBF network structure and improving classification performance

IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

Quantified Score

Hi-index	0.00

Visualization

Abstract

The major problems of currently used approaches for extracting symbolic rules from trained neural networks are analyzed. The lack of efficient heuristic information is the fundamental reason that causes the low effectiveness of currently used approaches. In this paper, a new measurement method of the classification power of attributes on the basis of differential information of the trained neural networks is proposed, which is suitable for both continuous attributes and discrete attributes. Based on this new measurement method, a new approach for rule extraction from trained neural networks and classification problems with continuous attributes is proposed. The performance of the new approach is demonstrated by several computing cases. The experimental results prove that the approach proposed can improve the validity of the extracted rules remarkably compared with other rule extracting approaches, especially for the complicated classification problems.