Refining expert knowledge with an artificial neural network

International Journal on Artificial Intelligence Tools, 2001

Neurules are a kind of hybrid rules integrating neurocomputing and production rules. Each neurule is represented as an adaline unit. Thus, the corresponding neurule base consists of a number of autonomous adaline units (neurules). Due to this fact, a modular and natural knowledge base is constructed, in contrast to existing connectionist knowledge bases. In this paper, we present a method for generating neurules from empirical data. To overcome the difficulty of the adaline unit to classify non-separable training examples, the notion of 'closeness' between training examples is introduced. In case of a training failure, two subsets of 'close' examples are produced from the initial training set and a copy of the neurule for each subset is trained. Failure of training any copy, leads to production of further subsets as far as success is achieved.

Neurules are a kind of hybrid rules integrating neurocomputing and production rules. Each neurule is represented as an adaline unit. Thus, the corresponding rule base consists of a number of autonomous adaline units (neurules). Due to this fact, a modular and natural rule base is constructed, in contrast to existing connectionist rule bases. In this paper, we present a method for generating neurules from empirical data. We overcome the difficulty of the adaline unit to classify non-separable training examples by introducing the notion of 'closeness' between training examples and splitting each training set into subsets of 'close' examples.

Decision Support Systems, 1996

This research explores a new approach to integrate neural networks and expert systems. The integrated system combines the strength of rule-based semantic structure and the learning capability of connectionist architecture. In addition, the approach allows users to define logical operators that behave much similar to that of human expert decision making process. Neural Logic Network (NEULONET) is used as the underlying building unit. A rule-based shell like environment is developed. The shell is used to built a prototype expert decision support system for future bonds trading. The system also provides a way to behave like different experts responding to different users and giving advice according: to different environmental situations.

Proceedings of the 1984 annual conference of the ACM on The fifth generation challenge - ACM 84, 1984

Expert systems are generally described by a mixture of terms that confuse implementation language with knowledge structure and the search process. This confusion makes it difficult to analyze new problems and to derive a set of knowledge engineering principles. A rigorous, logical description of expert systems reveals that a small set of terms and relations can be used to describe many rule-based expert systems. In particular, one common method for solving problems is by classification---heuristically relating data abstractions to a preenumerated network of solutions. This model can be used as a framework for knowledge acquisition, particularly in the early stages for organizing the expert's vocabulary and decomposing problems.

IFIP Advances in Information and Communication Technology, 2010

This paper introduces a tool, namely ACRES (Automatic CReator of Expert Systems), which can automatically produce rule-based expert systems as CLIPS scripts from a dataset containing knowledge about a problem domain in the form of a large number of cases. The rules are created via a simple systematic approach and make use of certainty factors (CFs). CFs of same conclusions can be combined either using the MYCIN method or a generalization of MY-CIN's method. This latter method requires calculation of some weights, based on a training dataset, via the use of a genetic algorithm. Creation of an expert system is outlined. Small scale experimental results comparing the above methods with each other and a neural network are finally presented.

Proceedings of 8th Mediterranean Electrotechnical Conference on Industrial Applications in Power Systems, Computer Science and Telecommunications (MELECON 96), 1996

This paper deals with a new methodology for developing an Expert System (ES). It has the ability of learning to extract knowledge from a poor knowledge base using the learning by example paradigms. The choice of a poor knowledge base was motivated by the fact that in this case it is easier to have consistence in putting together the several pieces of knowledge. So, the problems attached to knowledge elicitation are simplified.

The aim of this research is to develop a model for designing rule-based expert systems that uses the forward chaining method of inference. The striking aspect of this model is that the inference engine is based on a simple representation of rules and facts in relational database tables. Rules are decomposed and represented in tables at two levels, which allow the developing of expert systems in any programming language that supports SQL. The explanation facility uses tables containing the explanations of the result of each rule. The model proposed in this paper is based on a simple approach to represent facts and rules in relational database tables. The advantage of this model lies in focusing the design of rule-based expert systems toward knowledge representation in a relational database, reducing effort and programming difficulties.

2000

This paper reports on the design of an optimal knowledge base for integrated Artificial Neural Network (ANN) and Expert Systems (ES). In this system, an orthogonal plan is used to define an optimal set of examples to be taken from a problem domain. Then holistic judgments of experts on these examples will provide a training set for an ANN to serve as an initial knowledge base for the integrated system. Any counter-examples in generalization over the new cases will be added to the training set to retrain the network in order to enlarge its initial knowledge base.

Expert Systems with Applications, 2004

In this paper, we present an approach that integrates symbolic rules, neural networks and cases. To achieve it, we integrate a kind of hybrid rules, called neurules, with cases. Neurules integrate symbolic rules with the Adaline neural unit. In the integration, neurules are used to index cases representing their exceptions. In this way, the accuracy of the neurules is improved. On the other hand, due to neurule-based efficient inference mechanism, conclusions can be reached more efficiently. In addition, neurule-based inferences can be performed even if some of the inputs are unknown, in contrast to symbolic rule-based inferences. Furthermore, an existing symbolic rule-base with indexed exception cases can be converted into a neurule-base with corresponding indexed exception cases. Finally, empirical data can be used as a knowledge source, which facilitates knowledge acquisition. We also present a new high-level categorization of the approaches integrating rule-based and case-based reasoning. q

Expert Systems, 2002

Knowledge Acquisition, 1991

In this paper, we argue that techniques proposed for combining empirical and explanation-based learning methods can also be used to detect errors in rule-based expert systems, to isolate the blame for these errors to a small number of rules and suggest revisions to the rules to eliminate these errors. We demonstrate that FOCL, an extension to Quinlan's FOIL program, can learn relational concepts in spite of an incorrect domain theory (e.g. a knowledge base of an expert system that contains some erroneous rules). A prototype knowledge acquisition tool, KR-FOCL, has been constructed that utilizes a trace of FOCL to suggest revisions to a rule base.

Rule-extraction from trained neural networks has previously been used to generate propositional rule-sets. The extraction of "generic" rules or objects from trained feedforward networks is clearly desirable and sufficient for many applications. We present several approaches to generate a knowledge base that includes rules, facts and a is-a hierarchy that enables the greater explanatory capabilities by allowing the user interaction. The approaches are (1) construct two feedforward neural networks by cascade correlation algorithm [Fahlman & Lebiere, 1991] and tower algorithm ), extracts rules at the level of individual hidden and output units of both the networks by use of the decompositional rule-extraction method "LAP"; (2) cascade correlation and tower algorithm to train two different feedforward neural network, extracts rules that map inputs directly into outputs to generate the examples for each learning algorithm by the use of the pedagogical rule-extraction method "RuleVI" (3) constrained error back propagation to train a feedforward neural network, extract rules at the level of individual hidden and output units by use of the decompositional rule-extraction method "RULEX"; and use of the extracted symbolic rules to generate a connectionist knowledge base. Then the performance is demonstrated by a number of real-world applications.

2000

In this paper, two methods for extraction o f knowledge rules through Artificial Neural Networks, with continuous activation functions are presented. Those rules are extracted from neural networks previously trained and of the sensitivity factors obtained by the differentiation of a neural network. The rules can be used when analytic models of the physical processes lead to equations of difficult

Neural Networks, 1988

Applications usin~ expert systems are gaining in popularity. These systems apply expert knowledge about a particular field and logically progress through & given problem to arrive at an appropriate solution. Problems such as nuclear power plant monitoring, robot ma~Lipu~on, medical diagnosis are just a few examples of applications using expert systems. Present day expert systems are written on higher-level interpretive languages, st{ch as LISP and PRO~, running on fairly expeusi~;e computers. Even on these machines, the speed-performances of the expert systems are less than satisfactory as the number of rules increases. Hence these system are unsuitable for time-critical classification applications. Recently there is tremendous interest in artifical neural networks, which allows the parallel execution of competing hypothesis using massively parallel networks composed of many computational elements connected by links with variable weights. Implementation of expert systems using this ne.twork structure can improve the speed-performance of the system, and reduce complexity and time requirement for building the system. A methodology for building an expert system is developed. The methodology is based on the current concept of building expert systems, such as knowledge acquisition and representation, data collection, user interfacing. Additions] concepts of system learning(training), acceptability are also included. For demonstration purposes, a neural network learnin.g a]~orithm has been applied to develop a diagnostic expert system for patients with chest pain. The system co~ectly classifies four categories of ch~t pain: myo.c~, dial infarction, ischemic pain, non-ischemlc pain and non-cardiac pain. Simulatlon results will be presented in the paper and the conference to demonstratethe viability of the approach. The superiority of the neural networks approach will be proved by comparison of speed-performance o]~ neural networks approach vs. statistical approach. The system performance was comparable to the physicians.

2000

Contrary to the common opinion neural networks may be used for knowledge extraction. Recently a new methodology of logical rule extraction, optimization and application of rule-based systems has been described. C-MLP2LN algorithm, based on constrained multilayer perceptron network, is described here in details and the dynamics of a transition from neural to logical system illustrated. The algorithm handles real-valued features, determining appropriate linguistic variables or membership functions as a part of the rule extraction process. Initial rules are optimized exploring the tradeoff between accuracy/simplicity at the rule extraction stage and between reliability of rules and rejection rate at the optimization stage. Gaussian uncertainties of measurements are assumed during application of crisp logical rules, leading to "soft trapezoidal" membership functions and allowing to optimize the linguistic variables using gradient procedures. Comments are made on application of neural networks to knowledge discovery in benchmark and in real life problems.