Survey on: Fuzzy Verdict Reveal on Predefined Neural Network
Journal: International Journal of Science and Research (IJSR) (Vol.3, No. 11)Publication Date: 2014-11-05
Authors : Monika Prabhakar Wanjare; Y.B.Gurav;
Page : 2846-2848
Keywords : Classification; fuzzy min-max; hyperbox; machine learning; neural networks; neuron; supervised learning;
Abstract
In this paper, a supervised learning method, called as a multi-level fuzzy min-max neural network classifier (MLF), is described. MLF utilizes fundamental ideas of the fuzzy min-max (FMM) method in a multi-level structure to arrange patterns. This method uses separate classifiers with littler hyperboxes in distinctive levels to group the specimens that are found in overlapping areas. The last outcome of the network is shaped by consolidating the outcomes of these classifiers. MLF is fit for learning nonlinear limits with a solitary pass through the information. As per the acquired results, the MLF method, contrasted with the other FMM networks, has the most elevated execution and the least affectability to maximum size of the hyperbox parameter (), with the best preparation precision as in most of the cases.
Other Latest Articles
- Large-Scale Shell Model Investigation of Even-Even 66-76Ni Isotopes
- Novel Scheme to Segment Privacy and Securing Information to Brokering in Distributed Sharing
- A Survey Paper for Bug Localization
- A Novel Game-Theoretic Method for Detecting Malware in Delay-Tolerant Networks
- An Overview of Biometrics Methods?
Last modified: 2021-06-30 21:12:54