Neural networks have been successfully applied to solve a variety of application problems including classication and function approximation. They are especially useful as function approximators because they do not require prior knowledge of the input data distribution and they have been shown to be universal approximators. In many applications, it is desirable to extract knowledge that can explain how the problems are solved by the networks. Most existing approaches have focused on extracting symbolic rules for classication. Few methods have been devised to extract rules from trained neural networks for regression. This article presents an approach for extracting rules from trained neural networks for regression. Each rule in the extracted rule set corresponds to a subregion of the input space and a linear function involving the relevant input attributes of the data approximates the network output for all data samples in this subregion. Extensive experimental results on 32 benchmark data sets demonstrate the eectiveness of the proposed approach in generating accurate regression rules. Index terms: Regression, network pruning, rule extraction.
|
912
|
Practical Optimization
– Gill, Murray, et al.
- 1981
|
|
520
|
C4.5: Programs for
– Quinlan
- 1993
|
|
400
|
Fast learning in networks of locally-tuned processing units
– Moody, Darken
- 1989
|
|
180
|
A survey and critique of techniques for extracting rules from trained artificial neural networks
– Andrews, Diederich, et al.
- 1995
|
|
114
|
Learning with continuous classes
– Quinlan
- 1992
|
|
104
|
Dynamic node creation in backpropagation networks
– Ash
- 1989
|
|
60
|
The truth will come to light: Directions and challenges in extracting the knowledge embedded within trained artificial neural networks
– Tickle, Andrews, et al.
- 1998
|
|
42
|
Neural-Network Feature Selector
– Setiono, Liu
- 1997
|
|
41
|
Using relevance to reduce network size automatically
– Mozer, Smolensky
- 1989
|
|
41
|
Symbolic representation of neural networks
– Setiono, Liu
- 1996
|
|
38
|
Numerical Methods for Unconstrained Optimization and Nonlinear Equations
– Jr, E, et al.
- 1983
|
|
38
|
Induction of model trees for predicting continuous classes
– Wang, Witten
- 1997
|
|
34
|
Constructive algorithms for structure learning in feedforward ANNs for regression problems
– Kwok, Yeung
- 1997
|
|
31
|
Use of a quasi-Newton method in a feedforward neural network construction algorithm
– Setiono, Hui
- 1995
|
|
28
|
Extracting Rules from Neural Networks by Pruning and Hidden-unit Splitting
– Setiono
- 1997
|
|
26
|
A penalty function approach for pruning feedforward neural networks
– Setiono
- 1997
|
|
25
|
Objective functions for training new hidden units in constructive neural networks
– Kwok, Yeung
- 1997
|
|
19
|
Efficient algorithms for function approximation with piecewise linear sigmoidal networks
– Hush, Horne
- 1998
|
|
16
|
Automated Knowledge Acquisition
– Sestito, Dillon
- 1994
|
|
15
|
Determining Input Features for Multilayer Perceptron.” Neurocomputing Vol
– Belue, Bauer
- 1995
|
|
12
|
ANN-DT: An algorithm for extraction of decision trees from artificial neural networks
– Schmitz, Aldrich, et al.
- 1999
|
|
11
|
Combining regression trees and radial basis function networks
– Orr, Hallam, et al.
- 1999
|
|
11
|
Improved feature screening in feedforward neural networks
– Steppe, Bauer
- 1996
|
|
11
|
Exploring constructive cascade networks
– Treadgold, Gedeon
- 1999
|
|
10
|
Function approximation with spiked random networks
– Gelenbe, Mao, et al.
- 1999
|
|
10
|
Naive bayes for regression
– Frank, Trigg, et al.
|
|
9
|
FERNN: An algorithm for fast extraction of rules from neural networks
– Setiono, Leow
- 2000
|
|
7
|
Perturbation method for deleting redundant inputs of perceptron networks
– Zurada, Malinowski, et al.
- 1997
|
|
7
|
An empirical measure of element contribution in neural networks
– Mak, Blanning
- 1998
|
|
6
|
Integrated feature and architecture selection
– Steppe, Bauer, et al.
- 1996
|
|
6
|
The extraction of re rules from knowledge-based neural networks
– Towell, Shavlik
- 1993
|
|
6
|
GDS: Gradient descent generation of symbolic rules
– Blassig
- 1994
|
|
6
|
Generalized analytic rule extraction for feedforward neural networks
– Gupta, Park, et al.
- 1998
|
|
2
|
Symbolic Interpretation of Arti Neural Networks
– Taha
- 1996
|
|
1
|
Integrating arti neural networks with rule-based expert systems
– Yoon, Guimaraes, et al.
- 1994
|
|
