Hybrid Data Mining Based on Student Grade Clustering for Major Concentration Classification (Case Study: University Xyz in Serang, Banten)

The development of data mining is currently very much helping decision makers to determine the right decision for the problem at hand. Data mining is a technique used to analyze data by combining statistics, mathematics, and machine learning. In hybrid data mining, an important influencing factor is the process of generating different classifiers to form a general model. This study proposes to combine two data mining methods, clustering and classification. Thiscombination of data mining methods is called hybrid data mining. By utilizing the technique of grouping data based on similarities and differences.data errors can be minimized, thereby improving the accuracy of the classification model. Using hybrid data mining methods, where the is data clustered using k-medoid model and then classified, shows that 3 out of 5 models can be effectively applied. Based on the student learning data used as the test data, Hybrid data mining methods demonstrate higher accuracy rates, such as an increase from 74% to 92% for Naive Bayes, 86% to88% for Support Vector Machine, and 78% to 84% for K-Nearest Neighbor. so that concentration selection using hybrid data mining methods is more accurate.