Classification of Parkinson’s disease using Machine Learning Techniques

Abstract

Biomarkers derived from human speech could be used to investigate neurological illnesses such as paralysis agitans (PD). PD is a neurodegenerative disease that affects about one million people. Clinicians have previously relied on subjective grading systems to determine the severity of Parkinson's disease. Due to difficulties with control, vocalization can be used to detect and diagnose Parkinson's disease. As a result of technology improvements and the widespread usage of audio gathering equipment in everyday life, healthcare providers may profit from less expensive and more accurate diagnosis. Using a medical data dataset obtained from people with and without Parkinson's disease, we provide evidence to support this theory. Decision Tree, Logistic Regression, and Naive Bayes are examples of Machine Learning algorithms, as are Deep Learning algorithms like Recurrent Neural Networks.