Supervised Machine Learning Based Noninvasive Prediction of Atrial Flutter Mechanism from P-to-P Interval Variability under Imbalanced Dataset Conditions

<div>Specificity of proposed augmentation techniques by three classifiers (LDA, LOG, and SVM) for all feature subset lengths. Right <i>y</i>-axis indicates performance of the classifier using synthetic samples, and left <i>y</i>-axis shows the best oversample rate for a particular feature-length (red line).</div>

Computational Intelligence and Neuroscience

fig11

Figure 11

Figure 11: Supervised Machine Learning Based Noninvasive Prediction of Atrial Flutter Mechanism from P-to-P Interval Variability under Imbalanced Dataset Conditions 

Figure 11 | Supervised Machine Learning Based Noninvasive Prediction of Atrial Flutter Mechanism from P-to-P Interval Variability under Imbalanced Dataset Conditions 