Please use this identifier to cite or link to this item:
Title: El paquete methylearning y su aplicación Shiny: selección de marcadores y clasificación a partir de datos de metilación del DNA
Author: Piñeyro Valerio, David
Director: Adsuar Gómez, Antonio Jesús
Tutor: Marco Galindo, María Jesús
Others: Universitat Oberta de Catalunya
Keywords: feature selection
Issue Date: 6-Jun-2018
Publisher: Universitat Oberta de Catalunya
Abstract: The application of Machine Learning techniques, in the diagnostic and prevention context, is a hot topic nowadays. From all the biological data that can be used for this purpose, those generated by hight throughput techniques represent a great opportunity for the progress in medicine. Among them, epigenetic data has demonstrated its potential for the diagnostic in big social impact diseases, such as cancer and neurodegenerative diseases. However, the process of building a classifier model based on this kind of data is a hard process, which requires a fine balance between feature selection techniques, a necessity due to the dimensionality problem of the data, and the classification techniques. To provide a complete, robust and easy-to-use framework to assist the creation of classifiers based on epigenetic data, we created the methylearning package. Programmed in R, it is meant to assist in the process of feature selection and classification, making simple the exploration of multiple combinations and producing a wealth of information, numerical and graphical, to assess the performance of different algorithms. In addition, to open the use of the methylearning tool to a wider audience, a Shiny web application was also designed, which incorporates all the methylearning functionality, behind an easy-to-use graphical interface.
Language: Spanish
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
dpineyroTFM0618memoria.pdfMemoria del TFM1.42 MBAdobe PDFView/Open

This item is licensed under a Creative Commons License Creative Commons