Please use this identifier to cite or link to this item:
http://hdl.handle.net/10609/127626
Title: | Predicción de tiempo de fracaso de un tratamiento antirretroviral mediante algoritmos de machine learning de supervivencia |
Author: | Amado Bouza, Javier |
Tutor: | Perez-Alvarez, Nuria |
Others: | Maceira, Marc |
Abstract: | In this work, a study was carried out on the performance of machine learning algorithms applied to survival problems. The Lake database, obtained from a clinical trial named identically, was th chosen to carry on these protocol. In this clinical trial with antiretroviral treatments were tested. The problem of missing data is extremely common in longitudinal studies, wich include clinical trials. In addition, a comparison of the efficacy of the two antiretroviral treatments tested was made. Firstly, a data manegement process was carried out to prepare the data for a multiple imputation procedure, carried out using the MICE library. Once this imputation was made, 3 machine learning algorithms were applied to the imputed database (random survival forest, survival support vector machine, and boosting). In order to compare their performance, the predictions of the algorithms were used to calculate Harrell's C index. Finally, using the Wilcoxon test for paired samples, the efficacy of the antiretroviral treatments tested was compared. |
Keywords: | machine learning imputation VIH survival analysis |
Document type: | info:eu-repo/semantics/masterThesis |
Issue Date: | 12-Jan-2021 |
Publication license: | http://creativecommons.org/licenses/by-nc-nd/3.0/es/ |
Appears in Collections: | Trabajos finales de carrera, trabajos de investigación, etc. |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
javiamadoCodigo0121.Rmd | 23,98 kB | RMarkdown | View/Open | |
javiamadoGantt0121.gan | 15,86 kB | Gantt | View/Open | |
javiamadoVideo0121.mov | 160,31 MB | Video Quicktime | View/Open | |
javiamadoTFM0121memoria.pdf | 2,16 MB | Adobe PDF | View/Open | |
javiamadoTFM0121presentación.pdf | 863,51 kB | Adobe PDF | View/Open |
Share:
This item is licensed under a Creative Commons License