A Robust Ensemble Learning Framework for Early Detection and Classification of Parkinson's Disease Using Voice Features

Parkinson’s disease is a progressive neurodegenerative disorder that significantly impacts motor and vocal functions, making early and accurate diagnosis essential for effective treatment. This study presents an ensemble learning framework for the detection and classification of PD using biomedical voice features, offering a non-invasive and data-driven diagnostic approach. The proposed model combines multiple base classifiers—including Logistic Regression, K-Nearest Neighbors, and Decision Tree—through a soft voting mechanism to improve classification performance and reduce prediction variance. Experimental results demonstrate that the ensemble model outperforms individual classifiers, achieving an accuracy of 95%, with balanced precision, recall, and F1-score values of 91.5%. These results highlight the model’s robustness and generalizability, confirming its suitability for reliable Parkinson’s disease detection. The findings suggest that ensemble learning, when applied to voice-based biomedical data, can serve as an effective tool in clinical decision support systems for early PD diagnosis.