A comparative study of the svm and k-nn machine learning algorithms for the diagnosis of respiratory pathologies using pulmonary acoustic signals

Table 2 Performance outcome of the k-nn classifier for binary-normalised data

K value	Validation method	Classification accuracy (%)	Validation method	Classification accuracy (%)
1	Conventional validation	94.66 ± 0.56	Ten-fold cross-validation	96.65 ± 0.69
2		94.81 ± 0.42		96.35 ± 0.93
3		93.92 ± 0.68		95.06 ± 0.34
4		94.53 ± 0.47		94.41 ± 0.62
5		94.25 ± 0.94		95.02 ± 1.21
6		93.95 ± 1.15		94.41 ± 1.35
7		92.70 ± 1.27		93.20 ± 1.08
8		92.57 ± 0.85		94.44 ± 1.28
9		92.49 ± 0.63		93.32 ± 0.74
10		91.77 ± 0.77		92.17 ± 1.46

ISSN: 1471-2105