Automated gender‐Parkinson's disease detection at the same time via a hybrid deep model using human voice. (23rd August 2022)

• Record Type: Journal Article
• Authors: Kaya, Duygu
• Is Part Of: Concurrency and computation. Volume 34:Number 26(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-08-23
• Subjects: deep features -- feature selection algorithm -- k‐NN -- mRMR -- TQWT; Parallel processing (Electronic computers) -- Periodicals; Parallel computers -- Periodicals; 004.35
• DOI: 10.1002/cpe.7289
• Languages: English
• ISSNs: 1532-0626

Abstract

Summary: Gender and Parkinson disease (PD) identifications are critical parts to be noted from a given in human voice. Numerous artificial intelligence based methods have been proposed to detect gender and PD easily in literature. It is purposed to build an effective and a dependable simultaneously gender and PD recognition system based on feature extraction and feature selection methods in this study. First, CNN structure is used for obtaining deeper features from TQWT applied data and acoustic deep parameters are obtained by it. Later, these deep features are subjected to mRMR feature selection algorithm that increase the performance efficiency of the classifiers. As a result, the crucial features obtained by this hybrid structure and significant success rate 98.9% is obtained with the k‐NN classifier. Thus, gender and PD are detected at the same time. Also, this work is multiclass problem so, the other success parameters are calculated separately.