The straightforward mapping of a grammar onto a connectionist architecture is to make each grammar symbol correspond to a node and each rule correspond to a pattern of connections. The grammar then expresses the c̀ompetence' of the network. The (unsupervised) grammatical inference problem is therefore: how can a network learn to configure itself to reflect the syntactic structure in its input pa...

In this paper, we proposed a methods of implementation of intelligent controller for speed control of an induction motor using indirect vector control method has been analyzed in detail. Induction motor is used in many industrial applications of the total used electrical energy. This paper proposes a new control scheme based on artificial neural networks to obtain certain torque and speed opera...

In this paper we propose a new algorithm for rule extraction from a trained Multilayer Feedforward network. The algorithm is based on an interval arithmetic network inversion for particular target outputs. The types of rules extracted are N-dimensional intervals in the input space. We have performed experiments with four database and the results are very interesting. One rule extracted by the a...

Support Vector Machines (SVM) are believed to be as powerful as Artificial Neural Networks (ANN) in modeling complex problems while avoiding some of the drawbacks of the latter such as local minimæ or reliance on architecture. However, a question that remains to be answered is whether SVM users may expect improvements in the interpretability of their models, namely by using rule extraction meth...

Neural network technology has already been applied in a variety of domains with remarkable success. However, it has not been well utilized in data mining and knowledge discovery. In this paper, a general neural framework named NEUCRUM is proposed for classification rule mining. This paper also presents a possible implementation of NEUCRUM whose key components are a specific neural classifier na...

This paper demonstrates a mechanism whereby rules can be extracted from a feedforward neural network trained to characterize the money­ price relationship, defined as the relationship between the rate of growth of the money supply and inflation. Monthly Divisia component data is encoded and used to train a group of candidate connectionist architectures. One candidate is selected for rule extrac...

Bart Baesens • Rudy Setiono • Christophe Mues • Jan Vanthienen Department of Applied Economic Sciences, K. U. Leuven, Naamsestraat 69, B-3000 Leuven, Belgium Department of Information Systems, National University of Singapore, Kent Ridge, Singapore 119260, Republic of Singapore Department of Applied Economic Sciences, K. U. Leuven, Naamsestraat 69, B-3000 Leuven, Belgium Department of Applied E...

Breast cancer relapse prediction is an important step in the complex decision-making process of deciding the type of treatment to be applied to patients after surgery. Some non-linear models, like neural networks, have been successfully applied to this task but they suffer from the problem of extracting the underlying rules, and knowing how the methods operate can help to a better understanding...

The problem of inductive learning is hard, and| despite much work|no solution is in sight, from neural networks or other AI techniques. I suggest that inductive reasoning may be grounded in sensorimotor capacity. If an arti cial system to generalize in ways that we nd intelligent it should be appropriately embodied. This is illustrated with a network-controlled animat that learns n-parity by re...

We argue that despite being an actively researched area for nearly a decade, rule-extraction technology has not made as signiicant of an impact as it should have. A connuence of trends, however, has made the ability to extract comprehensible descriptions from complex learned models more important now than ever. We argue that rule-extraction methods can have a signii-cant impact in the overlappi...