An efficient dimensionality reduction method using filter-based feature selection and variational autoencoders on Parkinson's disease classification

• Published in: Biomedical signal processing and control Vol. 66; p. 102452
• Main Author: Gunduz, Hakan
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.04.2021
• Subjects: Dimensionality reduction; Fisher score; Multi-Kernel SVM; Parkinson's disease prediction; Relief; Variational autoencoder; Dimensionality reduction; Variational autoencoder; Fisher score; Multi-Kernel SVM; Parkinson's disease prediction; Relief
• ISSN: 1746-8094; 1746-8108
• DOI: 10.1016/j.bspc.2021.102452

•Our study proposes a PD diagnosis system based on different types of vocal features.•Relief and Fisher Score methods are combined with VAE to generate the deep features.•The efficacy of the proposed model are assesed with Multi-Kernel SVM classifier.•Deep Relief features result in an accuracy of 0.916 with 0.772 MCC rates.•All reduced feature sets have higher MCC rates than the features without selection. Parkinson's disease (Pd) is a progressive disease caused by the loss of brain cells and brings about speech and pronunciation defects during the early stages. This study revealed a Pd classification system based on vocal features extracted from the voice recordings of the individuals and proposed a hybrid dimensionality reduction methods to extract robust features. Proposed method took advantage of the prominent aspects of Variational Autoencoders (VAE) and filter-based feature selection models. Relief and Fisher Score were selected as filter-based methods for their effective performance in handling noisy data while VAE was used as a feature extractor due to the capability of preserving the regular latent space properties during the feature generation. In order to assess the effectiveness of the devised method, multi-kernel Support Vector Machines (SVM) classifier were trained with obtained deep feature representations. The combination of deep Relief features and SVM with multiple kernels distinguished Pd individuals from healthy subjects with an accuracy of 0.916 with 0.772 Matthews Correlation Coefficient (MCC) rates using only 30 features. Compared to results obtained without dimensionality reduction, proposed model provided approximately 9% and 22% improvements on accuracy and MCC rates, respectively. All experimental results showed that models trained with the deep features had higher accuracy and MCC rates with those trained with Fisher Score and Relief selected features. In addition, all models trained with reduced features had higher classification performance than the model without selection. It was also concluded that using multiple kernels in the SVM boosted the classification performance.