PARKINSON’S DISEASE PREDICTION USING MACHINE LEARNING

Parkinson’s Disease (PD) is a progressive neurodegenerative disorder that primarily impairs motor functions, speech, and cognitive ability. Detecting PD at an early stage is challenging because symptoms overlap with other neurological conditions and often manifest late, delaying medical intervention. This paper proposes a machine learning-based Parkinson’s Disease Prediction System that leverages clinical datasets and supervised algorithms for early detection. The system is developed using Support Vector Machine (SVM) and Random Forest classifiers, with Random Forest demonstrating superior accuracy and robustness. Clinical data is preprocessed, normalized, and feature-selected to optimize classification performance. The model is deployed as a Streamlit-based application that provides real-time predictions through an interactive interface. The system is further enhanced with a self-assessment module and an AI-powered chatbot using the Cohere API to improve user engagement and awareness. Experimental results reveal that Random Forest outperforms SVM in terms of accuracy, precision, recall, and F1-score, thereby establishing its reliability in predicting Parkinson’s Disease. This study demonstrates the practical role of machine learning in assisting healthcare professionals with early detection and supporting patients with accessible digital tools.