Skip to main content

Advertisement

Table 3 The average performances of kNN, QDA, and SVM classification methods in sleep stage classification on OSA group.

From: Sleep stage and obstructive apneaic epoch classification using single-lead ECG

Classification Sleep Stages Total
Method Wake NREM1 NREM2 NREM3 NREM4 REM Accuracy
kNN 817/1022 307/333 3124/4975 152/159 71/75 465/538 4957/7102
  = 80% = 92.3% = 62.8% = 95.5% = 94.6% = 86.4% = 69.8%
  CKI = 0.93 CKI = 0.93 CKI = 0.60 CKI = 0.94 CKI = 0.97 CKI = 0.95  
QDA 878/1022 317/333 3487/4975 154/159 72/75 492/538 5425/7102
  = 85.9% = 95.3% = 70.1% = 97.2% = 96.8% = 91.4% = 76.4%
  CKI = 0.97 CKI = 0.95 CKI = 0.65 CKI = 0.98 CKI = 0.98 CKI = 0.97  
SVM 883/1022 318/333 3517/4975 155/159 72/75 494/538 5461/7102
  = 86.4% = 95.5% = 70.7% = 97.6% = 97.0% = 91.9% = 76.9%
  CKI = 0.98 CKI = 0.98 CKI = 0.72 CKI = 0.98 CKI = 0.98 CKI = 0.97  
  1. The representation is similar to that of Table 2.