Identification of Parkinson’s Disease Based on Machine Learning Classifiers

Parkinson’s Disease (PD) is a progressive, enlightened, neurodegenerative and autoimmune untidiness that affects over 6.2 million individuals across the globe. This is a condition that mainly happens due to the loss of neurons that produce dopamine. As a result, it weakens the motor functionalities of the patient. Slowness in movements, tremors, shaking, stiffness are some of the primary symptoms of PD. As this disease is incurable, it is an important factor to classify at its early stage. Early detection can save people from more harm. Different approaches have been adopted by the researchers to identify PD from the normal people using the new age technologies i.e., Machine Learning (ML), Internet of Things (IoT), Wearable Body Area Network (WBAN), Deep Learning (DL), Reinforcement Learning (RL) etc. We have taken an open-source dataset to classify the same. The dataset contains vocal measurements of the subjects. We have taken different ML classifiers for classifying PD and obtained the models’ performance in terms of accuracy, precision, recall and F1 Score. An accuracy of 94.87% was obtained using Random Forest and K-Nearest Neighbor gave an accuracy of 93.56% after that. We were able to build the method of PD identification using ML classifiers and this article will be helpful for primordial detection and diagnosis of PD and also for tele-Medicare.