A Novel Metaheuristic Data Mining Algorithm for the Detection and Classification of Parkinson Disease

Objectives: Over the rapidly changing environment, the data mining techniques and its application are applied in the healthcare sector for medical diagnosis. The present study intends to provide a survey to gain the knowledge over current techniques of database discovery for the classification of Parkinson Disease. Methods: The study adopted a novel metaheuristic data mining algorithm for the detection and classification of Parkinson Disease were about 195 instances are selected for the investigation. In the initial phase the data underwent five phases, which includes training dataset, data pre-process, feature selection, classification and evaluation. However the research evaluated through performance measure tool, which consist of various techniques. This includes the confusion matrix, precision, recall and error rate. The confusion matrix is evaluated with various attributes like Specificity, Sensitivity, Accuracy and Positive and Negative predictive values. Findings: The study also performs a comparative study on five classification algorithms i.e. ABO, SCFW with KELM, FCM, ACO and PSO algorithms. This comparison results from confusion matrix of the selected algorithms which supports in identifying the specificity, sensitivity and accuracy of performance measures index showed that ABO algorithm is found to have best specificity, sensitivity and accuracy compared to all other algorithms, i.e. SCFW with KELM, FCM, PSO and ACO. In addition, the classifiers comparison results of the selected algorithms indicated that ‘ABO’ has the highest accuracy. Conclusion: In the present paper intended to estimate the efficiency and efficacy of the selected algorithm to best detect the Parkinson Dataset using various classifiers, as early detection of any kind of disease is an essential factor. The study reported that ABO algorithm has about have 97 percent accuracy in classifying and features filtering.