Parkinsons Disease Detection Using Machine Learning Models

Parkinson's disease is caused by the disruption of the brain cells that produce substances to allow brain cells to communicate with each other, called dopamine. The brain's dopamine-producing cells are what give movements control, adaptability, and fluidity. Parkinson's motor symptoms begin when 60 to 80% of these cells are destroyed because not enough dopamine is created. Researchers are working to find a way to identify the non-motor symptoms that arise early in the disease as early as possible in order to stop the disease's progression because it is believed that the disease starts many years before the motor (movement-related) symptoms. This research presents a Parkinson's disease diagnosis based on machine learning. Processes for feature selection and classification make up the suggested diagnosis technique. For the feature selection task, the Feature Importance and Recursive Feature Elimination approaches were taken into consideration. Parkinson's patients in the tests were categorized using KNN, Decision Tree, Random Forest, Naive Bayes, Linear Regression, Support Vector Machine, Gradient Boosting Classifier, AdaBoosting Classifier, and Neural Network. It was determined that Random Forest with recursive feature elimination outperformed the other techniques. In order to diagnose Parkinson's disease, the fewest possible vocal features yielded an 86.34% accuracy rate.