Table 10.

Performance results obtained with 2 classes (wheeze vs. other)—training: fixed duration; testing: fixed duration.

Classifiers	Accuracy	AUCWheez	F1Wheez	MCCWheez	F1Other	MCCOther
LDA_10MRMR	82.5 ± 0.2	83.0 ± 1.1	75.2 ± 0.5	74.6 ± 0.4	86.5 ± 0.2	84.1 ± 0.2
LDA_100MRMR	84.1 ± 0.0	88.6 ± 0.0	77.3 ± 0.1	77.2 ± 0.1	87.8 ± 0.1	85.8 ± 0.0
LDA_Full	83.3 ± 0.2	83.6 ± 0.1	78.1 ± 0.2	76.1 ± 0.2	86.5 ± 0.3	83.9 ± 0.5
SVMrbf_10MRMR	84.4 ± 1.3	87.3 ± 1.1	80.3 ± 2.3	78.1 ± 2.2	87.1 ± 0.8	84.4 ± 0.9
SVMrbf_100MRMR	87.2 ± 0.4	92.8 ± 1.4	84.1 ± 0.9	82.2 ± 0.7	89.3 ± 0.4	87.1 ± 0.5
SVMrbf_Full	88.6 ± 0.4	92.5 ± 1.1	86.1 ± 0.5	84.2 ± 0.6	90.3 ± 0.4	88.3 ± 0.4
RUSBoost_10MRMR	91.6 ± 1.2	96.2 ± 0.7	89.9 ± 1.5	88.6 ± 1.7	92.7 ± 1.1	91.3 ± 1.3
RUSBoost_100MRMR	91.0 ± 1.0	96.5 ± 0.6	89.7 ± 1.2	88.2 ± 1.4	91.9 ± 0.9	90.3 ± 1.1
RUSBoost_Full	93.6 ± 2.0	97.8 ± 0.8	92.5 ± 2.2	91.4 ± 2.6	94.4 ± 1.7	93.3 ± 2.1
CNN_dualInput	98.2 ± 0.4	98.1 ± 0.4	97.9 ± 0.5	96.4 ± 0.9	98.5 ± 0.4	96.4 ± 0.9
CNN_Spectrogram	98.6 ± 0.2	98.4 ± 0.2	98.3 ± 0.2	97.1 ± 0.4	98.8 ± 0.2	97.1 ± 0.4
CNN_melSpectrogram	98.3 ± 0.3	98.1 ± 0.2	97.9 ± 0.3	96.4 ± 0.6	98.5 ± 0.2	96.4 ± 0.6