Detecting Alzheimer's Disease by the Decision Tree Methods Based on Particle Swarm Optimization

Abstract

In this study aims to determine the classification of Alzheimer's disease, this disease is a dangerous disease that can eliminate memory loss and can even result in a loss of ability to remember. For this reason, early detection of this disease is needed so that it can prepare for medical treatment. In this study the proposed method is to compare several decision tree methods with featureor attribute selection using the Particle Swarm Optimization (PSO) algorithm with the Alzheimer OASIS 2 dataset: Longitudinal Data from kaggle.com. The results of experiments with ten-fold cross validation, by testing the decision tree algorithm before the feature or attribute selection is performed, the highest accuracy value is obtained from the random forest algorithm with a value of 91.15%. The feature selection process is carried out using the PSO algorithm and the experiment is repeated using the Decision tree, the PSO-based random forest algorithm has the highest accuracy value of93.56% with a kappa value of 0.884. Feature or attribute selection using the PSO algorithm is proven to be able to improve the accuracy of the decision tree algorithm, and is included in the algorithm with a very good range of values.