Abstract
A novel fuzzy rough granular neural network (NFRGNN) based on the multilayer perceptron using backpropagation algorithm is described for fuzzy classification of patterns. We provide a development strategy of knowledge extraction from data using fuzzy rough set theoretic techniques. Extracted knowledge is then encoded into the network in the form of initial weights. The granular input vector is presented to the network while the target vector is provided in terms of membership values and zeros. The superiority of NFRGNN is demonstrated on several real life data sets.
Highlights
Granular Computing (GC) is a new informationprocessing paradigm being developed in the past few years
Initial weights are determined among the nodes of all layers in the network by fuzzy rough sets based on fuzzy reflexive relation; 2
We have presented the design of a novel granular neural network architecture by integrating fuzzy rough sets with multilayer perceptron using back propagation algorithm
Summary
Granular Computing (GC) is a new informationprocessing paradigm being developed in the past few years. It recognizes that precision is an example in modeling and controlling complex systems. To enable a system to tackle real-life ambiguous situations (Ref. 2) in a manner more akin to humans, one may incorporate the concept of granules into the neural networks, a biologically inspired computing paradigm. Zhang et al (Ref. 21) described granular neural networks using fuzzy sets as their formalism and an evolutionary training algorithm. As part of determining the initial weights, Banerjee et al (Ref. 5) described a knowledge based network, where knowledge is extracted from data in the form of decision rules using rough set theoretic techniques. The main purpose of fuzzy and rough hybridization is to provide high degree of flexibility (Ref. 9), robust solutions (Ref. 10), and handling uncertainty (Ref. 11)
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: International Journal of Computational Intelligence Systems
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
