Comparison of SVM and ANFIS for Snore Related Sounds Classification by Using the Largest Lyapunov Exponent and Entropy

Table 6

Snore sound classification studies and their accuracy results.

Study	Duckitt et al. [17]	Cavusoglu et al. [18]	Karunajeewa et al. [19]	Yadollahi and Moussavi [20]	Our method
Sound types	Ambient sound	Ambient sound	Ambient sound	Ambient and tracheal sound	Ambient sound
Classes	Snoring and other sounds (silence, breathing, and other types of sounds); snore detection	Snore/nonsnore	Snore, breathing, and silence	Snore and breathing	Snore, breathing, and silence
Features	39-dimensional feature vector of energy and MFCC	Spectral energy distributions	Zero-crossings and signal’s energy	Zero-crossings signal’s energy, and first formant	The largest Lyapunov exponent (LLE) and entropy
Classifier	HMM	Linear regression	Minimum-probability-of-error decision rule	FLD	M-SVMs and ANFIS
Accuracy	82–89% snore sensitivity	86.8% snore sensitivity	90.74% total sensitivity	93.2% for ambient sound total accuracy	In Exp. I: 91.61% (SVMs), 86.75% (ANFIS) total accuracies; and 91.49% (SVMs), 79.31% (ANFIS) snore sensitivities