Un modelo para la extracción de conocimiento en base de datos, mediante cómputo evolutivo, automátas finitos deterministas y reglas difusas
Cargando...
Fecha
2019
Autores
Arcón Pineda, Aidys
Título de la revista
ISSN de la revista
Título del volumen
Editor
Ediciones Universidad Simón Bolívar
Facultad de Ingenierías
Facultad de Ingenierías
Resumen
El objetivo de esta tesis es el desarrollo de un método para la extracción de conocimiento en grandes bases de datos. Hoy en día se genera a través de la tecnología grandes cantidades de datos, tanto en volumen como en la dimensionalidad de sus variables. En este sentido, es necesario resaltar que la manipulación de los datos con un número elevado de variables presenta un problema para las técnicas tradicionales. Por otra parte, el conjunto de soluciones alternativas es tan elevado que la obtención de un óptimo global es inalcanzable en un tiempo razonable. Por tanto, es imprescindible emplear técnicas basadas en meta heurísticas que han demostrado ser una alternativa en la solución de problemas de gran complejidad por su aplicabilidad y eficiencia.
El presente trabajo propone un nuevo modelo EKCEAD (A model for the extraction of knowledge in Databases through evolutionary computation, finite deterministic automata and fuzzy rules), para la extracción de conocimiento en grandes bases de datos en forma de reglas difusas mediante el desarrollo de varias herramientas orientadas a procesos de búsqueda y optimización que utiliza las ventajas de autómatas finitos en la obtención de la mejor solución. El cual se prueba con casos reales comparándolo con otros métodos similares referenciados en la literatura, los resultados obtenidos han sido validados mediante el uso de pruebas estadísticas no paramétricas, que muestran un buen desempeño en términos de precisión e interpretabilidad.
Los resultados reflejan la bondad del modelo propuesto permitiendo que sea recomendado en la extracción de conocimiento en base de datos e incentivando su uso en futuras investigaciones.
The objective of this thesis is the development of a method for extracting knowledge in large databases. Today, large amounts of data are generated through technology, both in volume and in the dimensionality of their variables. In this sense, it should be emphasized that the manipulation of data with a large number of variables presents a problem for traditional techniques. On the other hand, the set of alternative solutions is high that obtaining an overall optimal is unattainable in a reasonable time. Therefore, it is essential to employ heuristic meta-based techniques that have proven to be an alternative in solving problems of great complexity because of their applicability and efficiency. This work proposes a new model EKCEAD (A model for the extraction of knowledge in Databases through evolutionary computation, finite deterministic automata and fuzzy rules), for knowledge extraction in large databases in the form of fuzzy rules by developing various tools oriented to search and optimization processes that uses the advantages of finite automata in obtaining the best solution. Which is tested against real-world cases against other similar methods referenced in the literature, the results obtained have been validated using nonparametric statistical tests, which show good performance in terms of accuracy and interpretability? The results reflect the goodness of the proposed model by allowing it to be recommended in the extraction of knowledge in database and incentivizing its use in future research.
The objective of this thesis is the development of a method for extracting knowledge in large databases. Today, large amounts of data are generated through technology, both in volume and in the dimensionality of their variables. In this sense, it should be emphasized that the manipulation of data with a large number of variables presents a problem for traditional techniques. On the other hand, the set of alternative solutions is high that obtaining an overall optimal is unattainable in a reasonable time. Therefore, it is essential to employ heuristic meta-based techniques that have proven to be an alternative in solving problems of great complexity because of their applicability and efficiency. This work proposes a new model EKCEAD (A model for the extraction of knowledge in Databases through evolutionary computation, finite deterministic automata and fuzzy rules), for knowledge extraction in large databases in the form of fuzzy rules by developing various tools oriented to search and optimization processes that uses the advantages of finite automata in obtaining the best solution. Which is tested against real-world cases against other similar methods referenced in the literature, the results obtained have been validated using nonparametric statistical tests, which show good performance in terms of accuracy and interpretability? The results reflect the goodness of the proposed model by allowing it to be recommended in the extraction of knowledge in database and incentivizing its use in future research.
Descripción
Palabras clave
Algoritmo evolutivo, Autómatas finitos deterministas, Lógica difusa, Extracción de conocimiento, Base de reglas difusas, Optimización, Evolutionary algorithm, Deterministic finite automatons, Fuzzy logic, Knowledge extraction, Fuzzy rule base, Optimization
