Abstract :

Parkinson’s disease (PD) is a progressive neurological disorder that affects motor and non-motor functions, making early diagnosis critical for effective treatment. Biomarker-based detection has gained significant attention for improving diagnostic accuracy. However, limited and imbalanced biomedical datasets restrict the performance of traditional machine learning models. Diffusion-based generative models have emerged as powerful tools for data augmentation and representation learning. This project proposes a diffusion-based generative modeling framework to enhance Parkinson’s biomarker detection. The generative model synthesizes realistic biomarker data to address data scarcity. Extracted features are classified using Random Forest (RF) and Support Vector Machine (SVM) algorithms. The hybrid approach improves robustness and generalization. The system evaluates performance using accuracy, precision, recall, and F1-score. Experimental results demonstrate improved detection accuracy