Harnessing the Power of Machine Learning for Parkinson's Disease Detection

Abstract

Parkinson's disease (PD) is a progressive neurodegenerative disorder affecting millions worldwide. Early detection of PD is crucial for effective treatment and management of the disease. Deep learning (DL) and machine learning (ML) have emerged as promising approaches for detecting PD. In this study, a comparative performance analysis is done for DL and ML applications based on speech signals. DL methods using convolutional neural networks (CNNs), long short-term memory (LSTM) networks, and ML methods employing random forest and the XGBoost model were trained and assessed. Performance of the models are evaluated using a variety of performance metrics, including accuracy, precision, recall, and F1-score. Results showed that the XGBoost model outperformed the DL models in terms of accuracy and F1 score, while the CNN and LSTM models achieved higher precision and recall. These findings suggest that XGBoost can be a useful tool for detecting PD based on speech signals, particularly in scenarios where interpretability and computational efficiency are important.