Phonocardiography Signal Classification by Applying Feature Space Transformations

摘要

Four feature extraction methods recently proposed for feature reduction and classification of high dimensional data (especially hyperspectral images), are assessed for applying to cardiac phonocardiography (PCG) signal for the first time in this paper. Clustering based feature extraction (CBFE), feature extraction using attraction points (FEUAP), feature space discriminant analysis (FSDA) and double discriminant embedding (DDE) methods are used for feature extraction from the PCG signal before applying the nearest neighbor classifier. The experiments demonstrate the good efficiency of these methods with respect to some other feature extraction ones used for PCG classification.