A Fuzzy Neural Network Approach Based on Dirichlet Tesselations for Nearest Neighbor Classi cation of Patterns
نویسندگان
چکیده
A neural network classiier using fuzzy set representation of pattern classes is presented. Network construction and learning is performed incrementally in a single pass by building an aggregate of space-lling regions that constitutes a simpliied variant of the construction known as Dirichlet tesselation (or Voronoi diagram). Each region is delimited by a set of hyperplanes and is endowed by a fuzzy membership function that forms the basis of learning and recall. Experimental results concerning diicult recognition problems show that the proposed approach is very successful in applying fuzzy sets to pattern classiication.
منابع مشابه
A Novel Fuzzy Based Method for Heart Rate Variability Prediction
Abstract In this paper, a novel technique based on fuzzy method is presented for chaotic nonlinear time series prediction. Fuzzy approach with the gradient learning algorithm and methods constitutes the main components of this method. This learning process in this method is similar to conventional gradient descent learning process, except that the input patterns and parameters are stored in mem...
متن کاملNearest Neighbor Classi cation from Multiple Feature Subsets
Combining multiple classi ers is an e ective technique for improving accuracy. There are many general combining algorithms, such as Bagging, Boosting, or Error Correcting Output Coding, that signi cantly improve classi ers like decision trees, rule learners, or neural networks. Unfortunately, these combining methods do not improve the nearest neighbor classi er. In this paper, we present MFS, a...
متن کاملIdentification of selected monogeneans using image processing, artificial neural network and K-nearest neighbor
Abstract Over the last two decades, improvements in developing computational tools made significant contributions to the classification of biological specimens` images to their correspondence species. These days, identification of biological species is much easier for taxonomist and even non-taxonomists due to the development of automated computer techniques and systems. In this study, we d...
متن کاملClassi¢cation of organic solvents and modelling of their physico-chemical properties by chemometric methods using different sets of molecular descriptors
Different sets of molecular descriptors using the k-nearest neighbor classi¢cation method were used to make a general classi¢cation of 152 organic solvents, the classi¢cation being further improved by performing the counterpropagation arti¢cial neural network. An extensive investigation was made of the physico-chemical properties of 152 solvents in a search for quantitative structure^property r...
متن کاملPattern Recognition using Optimized Prototypes
Hong Yan Department of Electrical Engineering, University of Sydney, NSW 2006, Australia Phone: +61 2 692-3515, fax: +61 2 692-3847, e-mail: [email protected] Abstract In this paper we present a method for designing optimized nearest neighbor classi ers. In this method, we produce initial prototypes for classi cation based on self-organization and optimize the prototypes using a multi-layer neura...
متن کامل