Please use this identifier to cite or link to this item: http://hdl.handle.net/10609/146347
Title: Comparison of machine learning algorithms in prediction of patients' survival using a health record database
Author: Jolis Orriols, Núria
Tutor: Perez-Alvarez, Nuria  
Others: Ventura, Carles  
Abstract: Machine learning is an emerging area that creates computer systems that by using algorithms and statistical models are capable of learning from existing data and making inferences to new data. The development of machine learning models has been a tool for working with large databases such as electronic health records to improve healthcare quality, efficiency, clinical research and capture billing data. The main objective of this TFM has been to infer on patients¿ survival prediction using an electronic health record and through the implementation of three machine learning classification algorithms. To do this, a basic protocol for beginners in machine learning has been developed which consists of six steps: (1) an exploratory analysis of the data with univariate and bivariate statistical analysis, (2) cleaning and curing of the data so that it can be analyzed, (3) multivariate analysis to know the relationship of predictive variables and their interaction with the response variable, (4) application of 3 of the most common machine learning classification models, (5) validation using k-fold cross-validation technique, (6) finally an evaluation and comparison of the generated models by means of some parameters such as balanced accuracy and AUC.
Keywords: machine learning
heart failure
predictive models
Document type: info:eu-repo/semantics/masterThesis
Issue Date: Jun-2022
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 SizeFormat 
njolisFMDP0622report.pdfReport of TFM1,7 MBAdobe PDFThumbnail
View/Open