Comparison of SVM and k-NN classifiers in the estimation of the state of the arteriovenous fistula problem
Marcin Grochowina, Lucyna Leniowska
DOI: http://dx.doi.org/10.15439/2015F194
Citation: Proceedings of the 2015 Federated Conference on Computer Science and Information Systems, M. Ganzha, L. Maciaszek, M. Paprzycki (eds). ACSIS, Vol. 5, pages 249–254 (2015)
Abstract. The publication presents a concise report on the comparison of the classifiers k-NN and SVM in the case of a fuzzy classification of the arterio-venous fistula based on audio recordings. What has been used in the studies are the acoustic signals taken from both healthy patients as well as those diagnosed with the narrowing of a fistula in a mild and major degree of stenosis. In the publication there have been selected two features, each presenting one- time and frequency domain, which enable a quite clear depiction of the classification result. The aim of the study is to develope a solution enabling the detection of fistula's pathologies at an early stage.