In order to improve the assessment of speech disorders in the context of Parkinson's disease, we have collected 34 sustained vowel / a /, from 34 subjects including 17 Parkinsonian patients. We subsequently extracted from 1 to 20 coefficients of the Mel-Frequency Cepstral Coefficients (MFCCs) from each subject. The frames were compressed using vector quantization, with six Codebook sizes. We used a leave-one-subject-out (LOSO) validation scheme and the Support Vector Machines (SVM) classifier with its different types of kernels, (i.e.; RBF, Linear and polynomial). After viewing the obtained results, we proceeded to a bench of 100 trials. The best average result obtained was 82% using the codebook size of 1.
Mendeley saves you time finding and organizing research
Choose a citation style from the tabs below