Advancing Parkinson’s Disease Detection: Integrating Machine Learning with Enhanced Feature Selection and Data Augmentation

Parkinson’s disease (PD) may be defined as one of the most devastating diseases of the nervous system, the victims of which number in the millions all over the world. This would mean early detection and, therefore, precise diagnosis, hence effective treatment and management for patients who are victims of Parkinson’s disease (PD). A number of the most encouraging findings reported in the literature are machine learning (ML) algorithms for the use of many metrics obtained from non-invasive and clinical tests in the diagnosis of Parkinson’s disease (PD). The study describes a thorough and efficient method for enhancing the use of machine learning (ML) algorithms in the diagnosis of Parkinson’s disease (PD), dealing with enhanced feature selection, data preprocessing, and the application of the SHAP method toward the evaluation of feature importance. Some of the advanced ML techniques elaborated on in subsequent sections include Random Forest, Decision Tree, AdaBoost, Extra Tree, and XGBoost among others. This line aims to address one of the aims of approaching problems and challenges represented by imbalanced datasets, as well as the identification of important objectives.