Early Diagnosis of Parkinson's Disease based on XGBoost Algorithm and Logistic Regression

Parkinson’s disease is a common neurodegenerative disease that affects nearly millions of people in the world. Parkinson's disease is often misdiagnosed or missed because its early symptoms are similar to aging. Starting from the early dysarthria of Parkinson's disease, this paper selects 22 of 195 samples for XGBoost model. Meanwhile, in order to avoid multicollinearity, Pearson correlation coefficient is used to reduce dimension to obtain 12 sound indicators for logistics model prediction. In this paper, the two models were comprehensively compared by means of confusion matrix thermal map, AIC value and sensitivity, and it was found that the former was 95% accurate but the sensitivity was high and the model stability was poor, while the latter was 90% accurate and the model goodness of fit was good. Therefore, the method discussed in this paper provides a new idea for the early identification of Parkinson's disease and points out the direction for further research.