Analysis of Different Modality of Data to Diagnose Parkinson's Disease Using Machine Learning and Deep Learning Approaches: A Review

Expert Systems · 2024

A review of machine learning and deep learning approaches for Parkinson's disease diagnosis across multiple data modalities — speech, gait, neuroimaging, and handwriting — surveying feature engineering, class-imbalance strategies (SMOTE / Borderline-SMOTE), and hyperparameter optimization techniques including the Grey Wolf Optimizer.