Neurodegenerative disorder diagnosis using support vector machine and Naive bayes algorithms

Abstract

The radical change of brain cells that causes dopamine, a component that allows brain cells to exchange information with one another, causes Parkinson's disease. Control, adaptation, and fluency of movement are all controlled by dopamine-producing cells in the brain. To reduce this production of dopamine, these cells should die at least 50%, resulting in Parkinson's motor symptoms. The diagnosis of Parkinson's disease using SVM and Navie bayes algorithms is presented in this paper. A feature selection and classification process is used in the proposed diagnosis method. In the experiments, the classification of diseased was done using Classification algorithms and Regression algorithms and Support Vector Machines. Our results compared Support Vector Machines with Feature Extraction outperformed the Naïve bayes. With the fewest number of features, 81.77 percent accuracy in Parkinson's diagnosis was achieved. This research work has preprocessed the dataset worked on Parkinson's Progression Markers Initiative (PPMI) and then used one of the classification methods, Support Vector Machine (SVM), to distinguish people with Parkinson's disease from healthy people. This article explained, how the ROC curve changes as the number of cross validation folds increases, as well as how the value of true positive and false positive rates changes.