Sistema para el análisis de técnicas descriptivas y regresión borrosa. Aplicaciones
Fecha
2011-07-04
Autores
Denoda Pérez, Lisset
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Central “Marta Abreu” de Las Villas
Resumen
En esta tesis se presentan los conceptos fundamentales de la teoría de conjuntos borrosos. Se definen los números borrosos triangulares, trapezoidales, así como sus operaciones fundamentales. Se muestran variantes para realizar el cálculo de medidas descriptivas borrosas como la media, moda, mediana y varianza. Se exponen los elementos fundamentales de la regresión borrosa haciendo énfasis en dos medidas de bondad de ajuste.
Se presentan los aspectos principales del análisis, diseño e implementación del software “efuzzy 1.0”.
Se explica el mecanismo que posee el Mathematica para intercambiar información con otros programas externos. Se presentan los pasos para conectar un programa en Java con el kernel del Mathematica mediante la biblioteca JLink. Finalmente se muestra la funcionalidad del Mathematica usada en el software “efuzzy 1.0”.
Se mostraron dos aplicaciones con datos reales. En el primer caso de estudio se realizó un tratamiento borroso para valorar el nivel de satisfacción de los clientes que asisten al área de cajeros en dos sucursales de BANDEC en Santa Clara. La modelación con números borrosos es buena ya que se reduce el grado de subjetividad del evaluador. En el segundo caso de estudio, se obtuvieron tres modelos de regresión borrosa aplicados a un problema de finanzas.
This thesis shows the principal concepts of the fuzzy set theory. Triangular fuzzy numbers and trapezoidal fuzzy numbers and their operations are defined. Several ways to calculated fuzzy descriptive statistics like mean, median, mode and variance, are shown. The fundamental elements of the fuzzy regression and two goodness of fit measures are also shown. The principal aspects related with the analysis, design and implementation of the software efuzzy 1.0 are presented. The Mathematica's algorithms to interchange information with external software are explained. Steps to connect a Java program with the kernel of Mathematica using JLink library are shown. Finally, the functions of Mathematica needed by effuzy were exposed. Two real data applications were presented. In the first study case, the satisfaction level of the clients of two banks of Santa Clara, was studied. The fuzzy number modelation is good because it reduces the subjectivity of the evaluator. In the second study case, three fuzzy regression models of the finances problem were obtained.
This thesis shows the principal concepts of the fuzzy set theory. Triangular fuzzy numbers and trapezoidal fuzzy numbers and their operations are defined. Several ways to calculated fuzzy descriptive statistics like mean, median, mode and variance, are shown. The fundamental elements of the fuzzy regression and two goodness of fit measures are also shown. The principal aspects related with the analysis, design and implementation of the software efuzzy 1.0 are presented. The Mathematica's algorithms to interchange information with external software are explained. Steps to connect a Java program with the kernel of Mathematica using JLink library are shown. Finally, the functions of Mathematica needed by effuzy were exposed. Two real data applications were presented. In the first study case, the satisfaction level of the clients of two banks of Santa Clara, was studied. The fuzzy number modelation is good because it reduces the subjectivity of the evaluator. In the second study case, three fuzzy regression models of the finances problem were obtained.
Descripción
Palabras clave
Diseño de Sistemas, Teoría de Conjuntos Borrosos, Regresión Borrosa, Medidas de Bondad de Ajuste, Software Efuzzy 1.0, Biblioteca JLink