Neural networks design: Rough set approach to continuous data (original) (raw)

Neural Networks and Rough Sets: A comparative study on data classification

The 2006 International …, 2006

This paper addresses a contrastive study between Neural Networks and Rough Sets on data classification. The experiments were carried out using the Iris database, of public domain, to evaluate the classification. The confusion matrix method was used to evaluate the performance of these classifiers. With these contrastive experiments, we investigated the capacity of each classifier for application in a potential application on knowledge extraction in databases. In this experiment the results indicate that the Neural Networks classifier, except SLP, presents significant superiority on Rough Sets classifiers.

Rough Set-Based Neuro-Fuzzy System

2009

This paper presents a novel hybrid intelligent system which synergizes the concept of knowledge reduction in rough set theory with the human-like reasoning style of fuzzy systems and the learning and connectionist structure of neural networks. The proposed rough set-based neuro-fuzzy system (RNFS) incorporates a wrapper-based feature selection method that employs the mutual information maximization scheme which selects attributes with high relevance and the concept of knowledge reduction in rough set theory which selects attributes with low redundancy. Experimental results show that the proposed RNFS utilizes less computational effort and yielded promising results on feature selection as well as classification accuracy.

Decision Rule Extraction from Trained Neural Networks Using Rough Sets

The ability of artificial neural networks to learn and generalize complex relationships from a collection of training examples has been established through numerous research studies in recent years. The knowledge acquired by neural networks, however, is considered incomprehensible and not transferable to other knowledge representation schemes such as expert or rule-based systems. Furthermore, the incomprehensibility of knowledge acquired by a neural network prevents users to gain better understanding of a classification task learned by the network. The aim of the present paper is to describe a method that can help to make the knowledge embedded in a trained neural network comprehensible, and thus transform neural networks into a powerful knowledge acquisition tool. Our method is based on rough sets, which offer a useful framework to reason about classification knowledge but lack in generalization capabilities. Unlike many existing methods that require training examples as well as th...

HYBRID EXPERT SYSTEM OF ROUGH SET AND NEURAL NETWORK

1999

The combination of neural network and expert system can accelerate the process of inference, and then rapidly produce associated facts and consequences. However, neural network still has some problems such as providing explanation facilities, managing the architecture of network and accelerating the training time. Thus to address these issues we develop a new method for preprocessing based on rough set and merge it with neural network and expert system. The resulting system is a hybrid expert system model called a Hybrid Rough Neural Expert System (HRNES).

Rough Set Approach to Unsupervised Neural Network based Pattern Classifier

2009

Abstract—Early Convergence, input feature space with minimal dimensions and good classification accuracy are always the most desired characteristics of an unsupervised neural network based pattern classifier. To achieve these, various approaches comprising of various soft computing tools can be used at different levels of implementations of such classifiers. Rough set is also one such tool, which can be used at either preprocessing level or learning level or architectural implementation level. Approaches using rough sets at first and the third levels are discussed here. Use of rough sets at the above stated levels result in dimensionality reduction of feature space through preprocessing and early convergence through rough neuron.

Rough Set Approach in Machine Learning: A Review

International Journal of Computer Applications, 2012

The Rough Set (RS) theory can be considered as a tool to reduce the input dimensionality and to deal with vagueness and uncertainty in datasets. Over the years, there has been a rapid growth in interest in rough set theory and its applications in artificial intelligence and cognitive sciences, especially in research areas such as machine learning, intelligent systems, inductive reasoning, pattern recognition, data preprocessing, knowledge discovery, decision analysis, and expert systems. This paper discusses the basic concepts of rough set theory and point out some rough set-based research directions and applications. The discussion also includes a review of rough set theory in various machine learning techniques like clustering, feature selection and rule induction.

New learning models for generating classification rules based on rough set approach

2000

Data sets, static or dynamic, are very important and useful for presenting real life features in different aspects of industry, medicine, economy, and others. Recently, different models were used to generate knowledge from vague and uncertain data sets such as induction decision tree, neural network, fuzz y logic, genetic algorithm, rough set theory, and others. All of these models take long time to learn for a huge and dynamic data set. Thus, the challenge is how to develop an efficient model that can decrease the learning time without affecting the quality of the generated classification rules. Huge information systems or data sets usually have some missing values due to unavailable data that affect the quality of the generated classification rules. Missing values lead to the difficulty of extracting useful information from that data set. Another challenge is how to solve the problem of missing data.

Hybrid rough sets/cellular neural networks approach to development of a decision making system

This paper describes a hybrid framework for this kind of switching circuit. In this framework, cellular neural networks (CNN) and rough sets are integrated into a hybrid system and used cooperatively during the system lifecycle. Rough sets and CNN were chosen for this application because they can discover patterns in ambiguous and imperfect data and provide tools for data and pattern analysis.

Design of rough neurons: Rough set foundation and Petri net model

2010

This paper introduces the design of rough neurons based on rough sets. Rough neurons instantiate approximate reasoning in assessing knowledge gleaned from input data. Each neuron constructs upper and lower approximations as an aid to classifying inputs. The particular form of rough neuron considered in this paper relies on what is known as a rough membership function in assessing the accuracy of a classification of input signals. The architecture of a rough neuron includes one or more input ports which filter inputs relative to selected bands of values and one or more output ports which produce measurements of the degree of overlap between an approximation set and a reference set of values in classifying neural stimuli. A class of Petri nets called rough Petri nets with guarded transitions is used to model a rough neuron. An application of rough neural computing is briefly considered in classifying the waveforms of power system faults. The contribution of this article is the presentation of a Petri net model which can be used to simulate and analyze rough neural computations.

Towards rough neural computing based on rough membership functions: Theory and application

2001

This paper introduces a neural network architecture based on rough sets and rough membership functions. The neurons of such networks instantiate approximate reasoning in assessing knowledge gleaned from input data. Each neuron constructs upper and lower approximations as an aid to classifying inputs. Rough neuron output has various forms. In this paper, rough neuron output results from the application of a rough membership function. A brief introduction to the basic concepts underlying rough membership neural networks is given.