Hybrid Weighted Ensemble Model for the Early Diagnosis of Parkinson’s Disease Using Voice Features

Parkinson’s Disease (PD) is the most prominent neurodegenerative disease after Alzheimer’s Disease which exhibits mild symptoms at initial stages with no cure. Early diagnosis of PD can delay the progression of the disease. The incorporation of Artificial Intelligence (AI) integrated with the non-invasive data capture method in the PD diagnosis system will provide robust and accurate diagnosis. The proposed system presents a novel Hybrid Weighted Ensemble Classifier Model for the early diagnosis PD system integrated with voice features, since voice-based symptoms are the most vital signs at an early stage of PD and also voice data acquisition is a non-invasive method. The proposed model is experimented with various combinations of weights, two different voting methods and with four different data sizes. The proposed model is tested with the largest voice feature dataset and evaluated with the evaluation metrics, namely Accuracy, F1 score, Precision and Recall. The results signify that the proposed “Hybrid Weighted Ensemble Model” has a better performance compared to the existing system found in the literature.