Enhancing Parkinson’s Disease Diagnosis Using Logistic Regression and KNN

Parkinson’s disease is a neurogenerative condition that affects motor control. The condition appears when dopamine-producing brain neurons, which are essential for controlling smooth muscle, begin to fail. Early diagnosis of Parkinson’s disease motor symptoms allows a patient to receive appropriate care at the appropriate moment. The most suitable machine learning method for Parkinson disease diagnosis is compared in this work between logistic regression and k-nearest neighbor (KNN). Voice measures are among the plethora of clinical and demographic data on the participants in the research dataset. Which composes of 31 people, 23 with Parkinson’s disease. Furthermore, voice attribute information used to detect the disease includes MDVP: Fo (Hz), MDVP: Fhi (Hz), himmer: DDA, and many more. Preliminary results show that both Logistic Regression and K-Nearest Neighbor can distinguish between people with Parkinson disease and healthy individuals, but K-Nearest neighbor algorithm outperformed Logistic Regression by achieving 86% accuracy, compared to the 85% accuracy attained by Logistic Regression. The findings suggest that Logistic Regression and KNN are viable tools for Parkinson disease prediction. The study shows that machine learning methods can be very useful for the early diagnosis of Parkinson disease.