Detlef Nauck

A neuro-fuzzy method to learn fuzzy classification rules from data

Abstract Neuro-fuzzy systems have recently gained a lot of interest in research and application. Neuro-fuzzy models as we understand them are fuzzy systems that use local learning strategies to learn fuzzy sets and fuzzy rules. Neuro-fuzzy techniques have been developed to support the development of e.g. fuzzy controllers and fuzzy classifiers. In this paper we discuss a learning method for fuzzy classification rules. The learning algorithm is a simple heuristics that is able to derive fuzzy rules from a set of training data very quickly, and tunes them by modifying parameters of membership functions. Our approach is based on NEFCLASS, a neuro-fuzzy model for pattern classification. We also discuss some results obtained by our software implementation of NEFCLASS, which is freely available on the Internet.

Obtaining interpretable fuzzy classification rules from medical data

For many application problems classifiers can be used to support a decision making process. In some domains-in areas like medicine especially-it is preferable not to use black box approaches. The user should be able to understand the classifier and to evaluate its results. Fuzzy rule based classifiers are especially suitable, because they consist of simple linguistically interpretable rules and do not have some of the drawbacks of symbolic or crisp rule based classifiers. Classifiers must often be created from data by a learning process, because there is not enough expert knowledge to determine their parameters completely. A simple and convenient way to learn fuzzy classifiers from data is provided by neuro-fuzzy approaches. In this paper we discuss extensions to the learning algorithms of neuro-fuzzy classification (NEFCLASS), a neuro-fuzzy approach for data analysis that we have presented before. We present interactive strategies for pruning rules and variables from a trained classifier to enhance its readability, and demonstrate our approach on a small example.

Neuro-fuzzy systems for function approximation

Abstract We present a neuro-fuzzy architecture for function approximation based on supervised learning. The learning algorithm is able to determine the structure and the parameters of a fuzzy system. The approach is an extension to our already published NEFCON and NEFCLASS models which are used for control or classification purposes. The proposed extended model, which we call NEFPROX, is more general and can be used for any application based on function approximation.

Real Time Business Intelligence for the Adaptive Enterprise

In todays competitive environment, analysing data to predict market trends and to improve enterprise performance is an essential business activity. However, it is becoming clear that business success requires such data analysis to be carried out in real-time, and that actions in response to analysis results must also be performed in real-time in order to meet the rapid change in demand from customers and regulators alike. This paper discusses issues and problems of current business intelligence systems, and then outlines our vision of future real-time business intelligence. We present a list of emerging technologies that are being developed within the research program of British Telecommunications plc (BT), which could contribute to the realisation of real-time business intelligence, in addition to some examples of applying these technologies to improve BTs systems and services

How the learning of rule weights affects the interpretability of fuzzy systems

Neuro-fuzzy systems have recently gained a lot of interest in research and applications. These are approaches that learn fuzzy systems from data. Many of them use rule weights for this task. In this paper we discuss the influence of rule weights on the interpretability of fuzzy systems. We show how rule weights can be equivalently replaced by modifications in the membership functions of a fuzzy system. We elucidate the effects rule weights have on a fuzzy rule base. Using our neuro-fuzzy model NEFCLASS we demonstrate the problems of using rule weights in a simple example, and we show that learning in fuzzy systems can be done without them.

A fuzzy neural network learning fuzzy control rules and membership functions by fuzzy error backpropagation

A kind of neural network architecture designed for control tasks is presented. It is called the fuzzy neural network. The structure of the network can be interpreted in terms of a fuzzy controller. It has a three-layered architecture and uses fuzzy sets as its weights. The fuzzy error backpropagation algorithm, a special learning algorithm inspired by the standard BP-procedure for multivariable neural networks, is able to learn the fuzzy sets. The extended version that is presented is also able to learn fuzzy-if-then rules by reducing the number of nodes in the hidden layer of the network. The network does not learn from examples, but by evaluating a special fuzzy error measure.<<ETX>>

Generating classification rules with the neuro-fuzzy system NEFCLASS

Neuro-fuzzy systems have recently gained a lot of interest in research and application. In this paper, we discuss NEFCLASS (NEuro-Fuzzy CLASSification), a neuro-fuzzy approach for data analysis. We present new learning strategies to derive fuzzy classification rules from data, and show some results.

A neuro-fuzzy approach to obtain interpretable fuzzy systems for function approximation

Fuzzy systems can be used for function approximation based on a set of linguistic rules. We present a method to obtain the necessary parameters for such a fuzzy system by a neuro-fuzzy training method. The learning algorithm is able to determine the structure and the parameters of a fuzzy system from sample data. The approach is an extension to our already published NEFCON and NEFCLASS models which are used for control or classification purposes. The NEFPROX model, which is discussed in this paper is more general, and it can be used for any problem based on function approximation. We especially consider the problem to obtain interpretable fuzzy systems by learning.

Fuzzy data analysis with NEFCLASS

Fuzzy data analysis as we interpret it in this paper is the application of fuzzy systems to the analysis of crisp data. In this area, neuro-fuzzy systems play a very prominent role and are applied to a variety of data analysis problems like classification, function approximation or time series prediction. Fuzzy data analysis in general and neuro-fuzzy methods in particular make it easy to strike a balance between accuracy and interpretability. This is an interesting feature for intelligent data analysis and shall be discussed in this paper. We interpret data analysis as a process that is exploratory to some extent. In order for neuro-fuzzy learning to support this aspect we require fast and simple learning algorithms that result in small rule bases, which can be interpreted easily. The goal is to obtain simple intuitive models for interpretation and prediction. We show how the current version of the NEFCLASS structure learning algorithms support this requirement.

DESIGNING NEURO-FUZZY SYSTEMS THROUGH BACKPROPAGATION

The goal of neuro-fuzzy combinations is to obtain adaptive systems that can use prior knowledge, and can be interpreted by means of linguistic rules. Neuro-fuzzy models can be divided into cooperative models, which use neural networks to determine fuzzy system parameters, and hybrid models which create a new architecture using concepts from both worlds. Besides this, there are concurrent neural/fuzzy models that use neural networks and fuzzy systems separately. Most approaches adapt the backpropagation learning rule [33] for neural networks. Some of these systems are discussed in the following pages.