Parkinson's disease (PD) is one of the major public health problems in theworld. It is a well-known fact that around one million people suffer fromParkinson's disease in the United States whereas the number of people sufferingfrom Parkinson's disease worldwide is around 5 million. Thus, it is importantto predict Parkinson's disease in early stages so that early plan for thenecessary treatment can be made. People are mostly familiar with the motorsymptoms of Parkinson's disease, however, an increasing amount of research isbeing done to predict the Parkinson's disease from non-motor symptoms thatprecede the motor ones. If an early and reliable prediction is possible then apatient can get a proper treatment at the right time. Nonmotor symptomsconsidered are Rapid Eye Movement (REM) sleep Behaviour Disorder (RBD) andolfactory loss. Developing machine learning models that can help us inpredicting the disease can play a vital role in early prediction. In thispaper, we extend a work which used the non-motor features such as RBD andolfactory loss. Along with this the extended work also uses importantbiomarkers. In this paper, we try to model this classifier using differentmachine learning models that have not been used before. We developed automateddiagnostic models using Multilayer Perceptron, BayesNet, Random Forest andBoosted Logistic Regression. It has been observed that Boosted LogisticRegression provides the best performance with an impressive accuracy of 97.159% and the area under the ROC curve was 98.9%. Thus, it is concluded that thesemodels can be used for early prediction of Parkinson's disease.
展开▼
