Abstract
In the medical field, a number of data-mining methods and algorithms have been applied to help the decision-making process extract meaningful information from medical data. The goal of data mining is to establish efficient analysis and capture the hidden features of medical data. This paper is aims to find out the efficient model to identified the Parkinson disease people. Some experiments will be run to classify healthy people from those with Parkinson's disease. Data are recorded for 14 patients and 15 healthy individuals. A comparative study of the performance of multilayer neural network, support vector machine and decisiontree classifiers. the features derived from temporal domain and frequency domain will be used to train each classifier. The performances of the classifiers are evaluated using five metrics: classification accuracy, sensitivity, specificity and area under the receiver operating characteristic curve. The best classification accuracy achieved by multi layer neural network is 91.18% using the extracted features and clinical information.