Genetic programming application for features selection in task of arterial hypertension classification

Kublanov Vladimir, Dolganov Anton, Hugo Filipe Silveira Gamboa

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

6 Citations (Scopus)

Abstract

The paper investigates the possibilities of the genetic programming approach in task of arterial hypertension patients diagnosing. For this purpose, the 3-stage functional clinical study involving the tilt test was performed on two groups: relatively healthy volunteers and patients suffering from the arterial hypertension of II-III degree. The study was focused on the analysis of the 64 features of heart rate variability signals, evaluated by the time-domain, frequency-domain (Fourier and wavelet) and nonlinear methods. Performance of different machine learning approaches was compared: Discriminant Analysis, Nearest Neighbors, Decision Trees and Naive Bayes. All calculations were performed in the in-house software written on Python. The results of genetic programming application show the significant improvement of the classification accuracy over the previously obtained results of search on the non-correlated features space.

Original languageEnglish
Title of host publicationProceedings - 2017 International Multi-Conference on Engineering, Computer and Information Sciences, SIBIRCON 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages561-565
Number of pages5
ISBN (Electronic)9781538615966
DOIs
Publication statusPublished - 14 Nov 2017
Event2017 International Multi-Conference on Engineering, Computer and Information Sciences, SIBIRCON 2017 - Novosibirsk, Russian Federation
Duration: 18 Sep 201722 Sep 2017

Conference

Conference2017 International Multi-Conference on Engineering, Computer and Information Sciences, SIBIRCON 2017
CountryRussian Federation
CityNovosibirsk
Period18/09/1722/09/17

Keywords

  • Classification
  • Functional studies
  • Genetic programming
  • Heart rate variability
  • Machine learning
  • Tilt-test

Fingerprint

Dive into the research topics of 'Genetic programming application for features selection in task of arterial hypertension classification'. Together they form a unique fingerprint.

Cite this