Identifying Academic Performance Patterns Among PTIK Students Using K-Means Clustering

Abstract

This study explores the identification of academic performance patterns among students in the Informatics and Computer Engineering Education Study Program (PTIK) at Sebelas Maret University, focusing on the 2022 cohort. Using the K-Means clustering method within the scope of Data Mining, this research analyzes student performance data across multiple course categories from the first to fourth semesters. Through the Elbow method, four optimal clusters were established, each representing distinctive patterns of academic achievement. The analysis was conducted using RapidMiner software to reveal nuanced insights into student learning outcomes. Cluster 1 consists of students with moderate achievements in most categories, with a particular strength in Multimedia. Cluster 2 includes students with generally lower academic performance but shows a relative strength in General Courses. Cluster 3 is composed of high-achieving students who excel across categories, particularly in Software Engineering (RPL), Multimedia, and Educational subjects, indicating well-rounded academic proficiency. Cluster 4 comprises students with notable strengths in Software Engineering and Computer Networking, yet demonstrates lower performance in certain specialized subjects. These findings highlight the potential to tailor educational programs to address the specific learning needs and strengths of each student group, facilitating more personalized and effective academic support.