International Journal of Scientific Research and Engineering Development

This study proposes and implements a framework for the objective, non-invasive screening of Parkinson's Disease (PD) leveraging advanced machine learning (ML) techniques. Traditional diagnostic methods are often subjective, timeconsuming, and prone to late-stage detection, which significantly hinders timely treatment. To address this critical gap, we utilize a publicly available dataset comprising biomedical vocal features (e.g., Jitter, Shimmer, and the Harmonicsto-Noise Ratio (HNR)), sourced from the UCI ML Repository. The core of the project involves training a robust classification model, specifically the Support Vector Machine (SVM). The model is engineered to accurately discriminate healthy individuals from PD patients. Crucially, the methodology incorporates feature standardization (using the formula x = (x - \mu) / \sigma) to ensure equitable contribution of all features during model training, mitigating bias from disparate measurement scales. system's predictive accuracy is rigorously validated on an independent test set using key performance indicators such as precision, recall, and the F1-score. Following this rigorous validation process, the highaccuracy model is finalized, saved using the Python pickle library, and then deployed as a live application within a lightweight Flask web framework. This complete implementation successfully creates a fast and universally accessible prediction tool designed to give healthcare professionals crucial assistance in achieving quicker and more reliable preliminary diagnoses. Overall, this project showcases a robust, end-to-end development pathway, effectively turning complex AI research into a practical, impactful tool for clinical pre-screening. This project presents a complete machine learning pipeline, from data preprocessing and model training to deployment, demonstrating how voice- based features can be effectively used for early Parkinson’s Disease risk assessment in a practical clinical support setting.The successful deployment into a user-friendly Flask application instantly transforms complex AI into a simple, noninvasive tool that any doctor can use, offering a beacon of hope where there was once only uncertainty, and giving people back crucial time to fight the disease on their own terms.

Dr.Manjunatha B.N, Ms.Priyanka, A.Sridevi, B.Harika, K.Harshitha,"Parkinson’s Disease Prediction Using ML" International Journal of Scientific Research and Engineering Development, V9(3): Page(441-449) May-June 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.