Title :
Feature set extension for heart rate variability analysis by using non-linear, statistical and geometric measures
Author :
Jovic, Alan ; Bogunovic, Nikola
Author_Institution :
Fac. of Electr. Eng. & Comput., Univ. of Zagreb, Zagreb, Croatia
Abstract :
The goal of this paper is to evaluate the application of a combination of heart rate variability features on successful classification of known heart disorders. We propose an extension over our previous work, which employs 11 features, both from non-linear and linear analysis of heart rate variability. The features were extracted from electrocardiogram recordings and analyzed in Weka system for data mining using several well-known classification algorithms: C4.5 decision tree, Bayesian network, random forest, and RIPPER rules. Significance of each feature is analyzed and the algorithms´ success rates are compared. The selected combination of features has a high classification potential.
Keywords :
Bayes methods; data mining; electrocardiography; feature extraction; medical signal processing; signal classification; Bayesian network; C4.5 decision tree; RIPPER rules; data mining; electrocardiogram recordings; feature set extension; heart disorder classification; heart rate variability analysis; heart rate variability features; random forest; Algorithm design and analysis; Biomedical monitoring; Classification algorithms; Data mining; Electrocardiography; Entropy; Feature extraction; Heart rate variability; Information analysis; Spatial databases; ECG classification; RIPPER; classification algorithms; geometric features; non-linear analysis; random forest;
Conference_Titel :
Information Technology Interfaces, 2009. ITI '09. Proceedings of the ITI 2009 31st International Conference on
Conference_Location :
Dubrovnik
Print_ISBN :
978-953-7138-15-8
DOI :
10.1109/ITI.2009.5196051
