Figure 3 Comparison of coefficient values for each class.
Related Figures (7)
Please cite this article as: Pawet Ptawiak, Novel Methodology of Cardiac Health Recognition Based on ECG Signals and Evolutionary-Neural System, Expert Systems With Applications (2017), doi: 10.1016/j.eswa.2017.09.022 Table 1: A description of the database with the selected 1000 ECG fragments along with the allocation of signals to the training and test sets for 10-fold cross validation method. rescaling: Figure 2: Comparison of the analyzed classes after applying the DFT and logarithm procedure: A — concentration within‘the classes: normal sinus rhythm (all fragments of the ECG signal, where the other colors represent signals from other patients); B — separation between all 17 classes (only the first fragments of the ECG signal for each class, where the other colors represent the signals from other classes). Table 5: A comparison ofthe results under the methods based on the subject-oriented validation scheme (da S. Luz et al., 2016).
