Evolutionary framework for DNA microarray cluster analysis

• Autores: José Antonio Castellanos Garzón
• Directores de la Tesis: Fernando Díaz Gómez (dir. tes.)
• Lectura: En la Universidad de Valladolid ( España ) en 2013
• Idioma: inglés
• Tribunal Calificador de la Tesis: Juan Manuel Corchado Rodríguez (presid.) , Carlos Javier Alonso González (secret.) , Florentino Fernández Riverola (voc.) , Gonzalo Gómez López (voc.) , Miguel Francisco de Almeida Pereira da Rocha (voc.)
• Enlaces
  • Tesis en acceso abierto en: UVADOC
• Resumen

  • This research work proposes an evolutionary framework where a method of hierarchical clustering represented by an evolutionary model, a set of cluster validation measures and a cluster visualization tool have been fused to create a suitable environment of knowledge discovery from DNA microarray data. On one hand, the clustering evolutionary model of our framework is a novel alternative that attempts to solve some of the problems faced by the existing clustering methods. On the other hand, our alternative of cluster visualization given by a tool couples new properties and visual components, allowing us to validate and analyze clustering results. It also allows a visual checking environment of the cluster validation measures. This way, the fusion of the clustering evolutionary model with the cluster visual model becomes our framework a novel application of data mining compared to the conventional methods of machine learning.

    In order to reach our proposal, we have focused our efforts on the combination of areas such as evolutionary computation, data mining and visual analytics to build the framework on the domain of gene expression data. Each of these areas provides techniques that play a major role for the analysis and resolution of the current challenges in Bioinformatics.

    As a final conclusion of our proposal, we state that, the fusion of both approaches to create the evolutionary framework has proved to be of great help in the understanding of the data and provided knowledge on the used clustering methods. Moreover, we have provided new visualization components which can also be used to validate the existing ones. The framework has proven that our evolutionary method performs well and that can find better solutions than the others, this is shown not only through validity measures but also, with result visualizations. This way, the visual part of our framework is able to display clustering results and validate both, cluster validity measures and more importantly, a visualization component with another. Consequently, the results have shown that our framework is a powerful tool for cluster analysis from DNA microarray data, not only for the introduced evolutionary model, but also for any other hierarchical clustering method that we want to add to the process of cluster analysis.