An Approach for Prototype Generation based on Similarity Relations for Problems of Classification

Authors

  • Yumilka B. Fernandez Hernandez Universidad de Camagüey
  • Rafael Bello Universidad Central de Las Villas
  • Yaima Filiberto Universidad de Camagüey
  • Mabel Frias Universidad de Camagüey
  • Lenniet Coello Blanco Universidad de Camagüey
  • Yaile Caballero Universidad de Camagüey

DOI:

https://doi.org/10.13053/cys-19-1-2053

Keywords:

Prototype generation, similarity relations, classification

Abstract

In this paper, a new method for solving classification problems based on prototypes is proposed. When using similarity relations for granulation of a universe, similarity classes are generated, and a prototype is constructed for each similarity class. Experimental results show that the proposed method has higher classification accuracy and a satisfactory reduction coefficient compared to other well-known methods, proving to be statistically superior in terms of classification’s precision. 

Author Biography

Yumilka B. Fernandez Hernandez, Universidad de Camagüey

Departamento de Computacion

Downloads

Published

2015-03-27