This project focuses on predicting Parkinson's clinical scores using voice features extracted from the ICEBERG Voice dataset. The dataset, which is not publicly available, includes clinical scores recorded from facial impressions and gait analysis.