<![CDATA[The process of monitoring and evaluating high school student leadership is deemed necessary because the level of student leadership is one of the prerequisites for high school students to face real challenges in the future. Data mining can be used to classify the level of leadership among high school students. The purpose of the research conducted in this case is to apply data mining using the K-NN, Decision Trees, and Naive Bayes models. This research is located in two different public high schools, namely SMA A as training data and SMA B as test data. This data was obtained in the same year, namely 2022. The data obtained were analyzed with the help of the Rapidminer application using K-NN, Decision Tree, and Naive Bayes. Student data that is processed is Basic Education Data (DAPODIK) in excel format. Before being analyzed, the text is processed first, namely tokenization, case folding, stop words, and details. The main goal of the steps above is also the main goal of this study to get the most accurate algorithm for classifying student leadership levels and knowing the results for comparison. The conclusion of this study is when measuring the performance of the three algorithms, the test results use confusion matrix validation. The K-NN algorithm was found to have the highest accuracy score compared to the Decision Tree and Naive Bayes. The accuracy value of the K-NN method using a dataset of high school students is 95.86%, the accuracy value of the Decision Tree algorithm is 94.65%, and the accuracy value of the Naïve Bayes algorithm is 79.55%.]]>
Copyrights © 2023
