Parkinson's Disease Identification from Speech Signals Using Machine Learning Models

Parkinson's disease (PD) is a common chronic neurodegenerative illness characterised by continuous nervous system degradation. This condition is more prevalent in the elderly. In Parkinson's, dopaminergic neurons die at an early stage, resulting in a progressive neurodegenerative condition. PD can cause a various symptom of non-motor and motor, including smell and speech. One of the problems that patients with Parkinson's may face is a pronunciation or having difficulty while speaking. As a result, early diagnosis is critical in minimising the potential effects of disease-related speech disorders. This journal intends to build a categorisation scheme for Parkinson's disease to distinguish between healthy individuals and PD sufferers and create a hybrid classifier by combining distinct machine learning models. For this journal, we have implemented Support Vector Machine (SVM), K-Nearest Neighbour (KNN), Random Forest classifier, and Logistic Regression ML techniques and acquired the classification report. The results showed that Random Forest has outperformed other ML techniques with 89.47% accuracy for the testing set.