Discrimination between normal and heart murmurs sound, based on statistical parameters extraction and classification

In this work, a new method for discrimination between normal and heart murmurs sound is presented. Statistical parameters, such as standard deviation (SD), are extracted from two datasets of heartbeats. Several classification technics, such as Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Naïve Bayes (NB), discriminative analysis, and classification tree, are used. Simulation results obtained from yielding methods are compared and discussed. The developed method (scheme) return good results from deferent dataset. Results obtained by using different classification methods versus two dataset are, significantly, accurate compared to the existing methods.