A boosting-based prediction model for disease progression

Abstract

With the intensification of the aging population in China, the number of patients with Parkinson's disease is showing a high growth trend. At present, the diagnosis of Parkinson's disease mainly relies on clinical assessment scales, but there are shortcomings such as poor accuracy, large scoring span and poor timeliness. Therefore, based on boosting, this paper uses a cross-validation algorithm to select the best model among 14 machine learning models and constructs the Multimode Parkinson model. We validate the performance of our model in predicting the extent of Parkinson's disease. Experimental results show that this method is superior to traditional algorithms and can be effectively applied to the treatment of Parkinson's disease.