Extracting rules from trained neural networks

Tsukimoto, Hiroshi

doi:10.1109/72.839008

Cited by 144 publications

(72 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Pruning techniques [8,9,10] begin by training a larger than necessary network and then reduce weights and neurons that are deemed redundant. Constructive algorithms present several significant advantages over pruning based algorithms including the ease of specification of the initial network topology, better economy in terms of training time and number of training examples, and potential for converging to a smaller network with superior generalization [11,12,13].…”

Section: Destructive Learningmentioning

confidence: 99%

Convergence Rate in a Destructive Neural Network With /Without Thresholds in The Output Layer

Mahmoud

2016

International Journal of Internet Education

View full text Add to dashboard Cite

Neural networks are a practicable solution for the extraction of accurate knowledge, where the data being mined can be so noisy, due to their relatively good tolerance to noisy and generalization ability and the performance of a neural network is directly related to its parameters and architecture. The degree of complexity of ANN increases exponentially as a factor of the numbers of input and hidden nodes. The complexity problem can be improved by constructing the structure of the network based on constructive learning and destructive learning. So, the objective of the network is to learn or to discover some involvement between input and output patterns to find the structure of the input patterns. The learning process is achieved through the modification of the connection weights between units. It is known as the network's topology which determines the network's final behavior. The goal of finding an optimal topology is to minimize an error function while conserving generalization capabilities. In our work, the destructive topology is used: first the algorithm trains a big size of neural network on the data, and then prunes it to increase its generalization capability while preserving its accuracy. The present paper introduces destructive neural network learning techniques and presents the analysis of the convergence rate of the error in a neural network with and without threshold in the output layer. The present paper introduces destructive neural network learning techniques and presents the analysis of the convergence rate of the error in a neural network with and without threshold.

show abstract

Section: Destructive Learningmentioning

confidence: 99%

Convergence Rate in a Destructive Neural Network With /Without Thresholds in The Output Layer

Mahmoud

2016

International Journal of Internet Education

View full text Add to dashboard Cite

show abstract

“…With data mining techniques, we can extract useful knowledge from training data to solve problems. Many methods have been proposed to generate fuzzy rules from training instances [3], [5], [6], [15], [17], [18] based on the fuzzy set theory [19]. In [6], we have presented a method for generating fuzzy rules from relational database systems for estimating null values.…”

mentioning

confidence: 99%

Generating weighted fuzzy rules from relational database systems for estimating null values using genetic algorithms

Chen

Huang

2003

IEEE Trans. Fuzzy Syst.

View full text Add to dashboard Cite

Abstract-In recent years, some methods have been proposed to estimate null values in relational database systems. However, the estimated accuracy of the existing methods are not good enough. In this paper, we present a new method to generate weighted fuzzy rules from relational database systems for estimating null values using genetic algorithms (GAs), where the attributes appearing in the antecedent part of generated fuzzy rules have different weights. After a predefined number of evolutions of the GA, the best chromosome contains the optimal weights of the attributes, and they can be translated into a set of rules to be used for estimating null values. The proposed method can get a higher average estimated accuracy rate than the methods we presented in two previous papers.

show abstract

“…This is due because neural networks have stored the knowledge in the weights linked to the connections, and therefore, it is difficult to explain the concepts in the weights from which neural networks elaborate the correct output. That is the reason Artificial Intelligence is performing some research about symbolic knowledge acquisition from a neural network [8,12,14]. Obtained rules of the neural network could give to knowledge engineering new points of view about the domain and new rules to interpret.…”

Section: Introductionmentioning

confidence: 99%

Data Mining with Enhanced Neural Networks-CMMSE

et al. 2013

View full text Add to dashboard Cite

This paper presents a new method to extract knowledge from existing data sets, that is, to extract symbolic rules using the weights of an Artificial Neural Network. The method has been applied to a neural network with special architecture named Enhanced Neural Network (ENN). This architecture improves the results that have been obtained with multilayer perceptron (MLP). The relationship among the knowledge stored in the weights, the performance of the network and the new implemented algorithm to acquire rules from the weights is explained. The method itself gives a model to follow in the knowledge acquisition with ENN.

show abstract

Extracting rules from trained neural networks

Cited by 144 publications

References 12 publications

Convergence Rate in a Destructive Neural Network With /Without Thresholds in The Output Layer

Convergence Rate in a Destructive Neural Network With /Without Thresholds in The Output Layer

Generating weighted fuzzy rules from relational database systems for estimating null values using genetic algorithms

Data Mining with Enhanced Neural Networks-CMMSE

Contact Info

Product

Resources

About