ABSTRACT
Parkinson's disease is a progressive neurological disorder characterized by motor and non-motor symptoms, often leading to significant impairment in quality of life. Early diagnosis is crucial for effective management and treatment. This project aims to develop a machine learning-based predictive model to aid in the early detection of Parkinson's disease using voice features. The dataset, obtained from Kaggle, includes various voice metrics such as vocal fundamental frequencies, measures of variation in frequency and amplitude, and noise-to-tonal component ratios. The implementation process involved several key steps: data preprocessing, feature selection, model selection, training, and validation. We explored multiple machine learning algorithms, including Logistic Regression, Support Vector Machine (SVM), Random Forest, and K-Nearest Neighbors (KNN). The Random Forest model was identified as the best-performing model based on cross-validation results, achieving an accuracy of 92%, precision of 90%, recall of 94%, and an F1-score of 92%.
Our findings demonstrate that the Random Forest model can effectively distinguish between Parkinson's and non-Parkinson's cases, with minimal false positives and false negatives. The system's high performance highlights its potential as a valuable tool for early diagnosis in clinical settings, potentially improving patient outcomes through timely intervention.
Future recommendations include expanding the dataset, exploring additional features, employing advanced machine learning techniques, and conducting real-world testing to validate the model's practical utility. This project underscores the promise of machine learning in enhancing diagnostic processes for Parkinson's disease and contributes to the growing body of research in the application of artificial intelligence in healthcare.
