Collection and Analysis of a Parkinson Speech Dataset with Multiple Types of Sound Recordings As the tecnological progress advances, usage of predictive telediagnosis and telemonitoring systems in the medical area is also increasing. These kind of systems have already been used for the diagnosis and monitoring of widespread diseases like tension (hyper/hipo) and diabetes. With the increasing interest to these systems, there has been an increased interest in speech pattern analysis applications of Parkinsonism for building predictive telediagnosis and telemonitoring models. For this purpose, we have collected a wide variety of voice samples, including sustained vowels, words, and sentences compiled from a set of speaking exercises for People with Parkison’s (PWP). There are two main issues in learning from such a dataset that consists of multiple speech recordings per subject: (i) how predictive these various types, e.g. sustained vowels vs. words, of voice samples are in Parkinson’s Disease (PD) diagnosis? (ii) how well the central tendency and dispersion metrics serve as representatives of all sample recordings of a subject? One of the aims of this thesis study is to investigate the performance of popular machine learning tools on different voice samples. Additionally the success of using the representations of the voice samples with the central tendency and dispersion metrics is also examined. This study showed that investigating our Parkinson dataset using well-known machine learning tools, as reported in the literature, sustained vowels are found to carry more PD-discriminative information. It is also found that rather than using each voice recording of each subject as an independent data sample, representing the samples of a subject with central tendency and dispersion metrics improves generalization of the predictive model.
