Systematic Literature Review: Perbandingan Algoritma Klasifikasi

Abstract

Over time, many data mining methods have been created and suggested to help in making decisions. Due to limited resources, this article only provides a systematic literature review in comparing the performance of the Naïve Bayes, Decision Tree, Nerral Network, Random Forest, and Support Vector Machine methods to find out which method is most effective in classifying and predicting. After conducting a literature study by taking articles from 2019 to 2023, 500 articles were obtained that used the Naive Bayes, Decision Tree, Neural Network, Random Forest, Support Vector Machine methods. Because there were so many articles obtained in the initial search, inclusion and exclusion criteria were created to sort out articles that were in accordance with this research. After carrying out the inclusion and exclusion criteria process, 243 articles were obtained and it was discovered that the topic that was discussed more often was prediction, which amounted to 122 articles and the remaining 121 articles discuss classification. In the field of prediction, the method most frequently used is Random Forest with a total of 45 articles and an average accuracy rate of 91.18%, while in the field of classification, the method most frequently used is Support Vector Machine with a total of 32 articles and an average accuracy of 88.85%.