Insight into an unsupervised two-step sparse transfer learning algorithm for speech diagnosis of Parkinson’s disease

Abstract

Speech diagnosis of Parkinson's disease (PD) as a non-invasive and simple diagnosis method is particularly worth exploring. However, the number of samples of speech-based PD is relatively small, and there exist discrepancies in the distribution between subjects. In order to solve the two problems, a novel unsupervised two-step sparse transfer learning is proposed in this paper to tackle with PD speech diagnosis. In the first step, convolution sparse coding with the coordinate selection of samples and features is designed to learn speech structure from the source domain to replenish sample information of the target domain. In the second step, joint local structure distribution alignment is designed to maintain the neighbor relationship between the respective samples of the training set and test set, and reduce the distribution difference between the two domains at the same time. Two representative public PD speech datasets and one real-world PD speech dataset were exploited to verify the proposed method on PD speech diagnosis. Experimental results demonstrate that each step of the proposed method has a positive effect on the PD speech classification results, and it also delivers superior performance over the existing relative methods.

Video surveillance using deep transfer learning and deep domain adaptation: Towards better generalization

“…• Accuracy, recall, precision, and specificity: to test the efficiency of a suggested method in ASR field, the classification accuracy, recall (sensitivity) or commonly known as true positive rate (TPR), precision (known also as positive predictive value), and specificity (commonly known as true negative rate (TNR)) are often used as assessment criteria for experiment results, such as in [56,30,57]. The accuracy rate is defined as the proportion of properly assessed samples to the total number of samples.…”

“…The proposed scheme presents an acceptable PD detection. In order to solve the scarcity of speech-based PD, and the existence of inconsistency in the distribution between subjects, a novel two-step unsupervised DTL algorithm called two-step sparse transfer learning (TSTL) [56] is proposed to deal with the above two mentioned problems. The method can assist in extracting useful information from large amounts of unlabeled speech data, aligning the distribution of the training and test sets, and preserving the original structure between samples all at the same time.…”