<![CDATA[The use of technology in the field of trade and sales is increasingly growing. Product information is an important role in building consumer trust when making purchasing decisions. Therefore, classification analysis is needed to help potential consumers in drawing conclusions. Classification analysis aims to conclude and identify data and classify its polarity. The Support Vector Machine (SVM) algorithm is widely used by many researchers for use in classification analysis. This algorithm was chosen because it can identify separate hyperplane to maximize the margin between 2 different classes. However, the Support Vector Machine (SVM) has deficiencies in parameter selection, so the selection of the Particle Swarm Optimization (PSO) feature is applied to improve accuracy. The results showed that implementation of the Support Vector Machine (SVM) has an accuracy value of 81.48% and an AUC value of 0.825, while optimization using Particle Swarm Optimization (PSO) has an accuracy value of 89.78% and an AUC of 0.902. The application of Particle Swarm Optimization (PSO) has been proven to improve the performance of the Support Vector Machine (SVM) algorithm.]]>
Copyrights © 2023
