A New Approach for Detection of Pathological Voice Disorders with Reduced Parameters

Voice data has demonstrated chaotic behavior in previous studies. Therefore, studying the linear properties alone does not yield successful results. This is valid for the examination of voice data as well. Therefore, conducting studies including chaotic features as well as existing technologies is inevitable. The main purpose of this study is to detect voice pathologies with fewer special features using new chaotic features. Both linear and nonlinear characteristics were used in this study. In this context, the largest Lyapunov exponents and entropy are preferred as chaotic properties because of their success in previous studies. Very few results with 100% accuracy were obtained in the experimental studies. In this study, multiple support vector machines (SVMs) were selected as a classifier because of their success in previous similar data types. Thus, the desired accuracy level was achieved using fewer features. Resultantly, the process complexity decreased and the system speed increased.