Understanding how students perform is crucial in educational data mining (EDM). By analyzing performance, academic institutions can identify important patterns, detect students who might be struggling, and develop effective support systems. This paper explores using clustering techniques to group students based on various performance indicators, including grades, attendance, engagement, and participation in extracurricular activities. Dividing students into distinct groups allows educators to better understand their learning behaviors, allocate resources more efficiently, and implement tailored intervention programs. This study provides a comparative evaluation of different clustering methods used for assessing student academic outcomes. We investigate the effectiveness of several well-known clustering algorithms, such as K-Means, DBSCAN, BIRCH, and Expectation Maximization (EM), in categorizing students based on their educational achievements. Our findings highlight the strengths and weaknesses of each method, offering valuable insights into their practical application within educational data mining. The significant growth in educational data has made advanced data mining techniques essential for extracting meaningful patterns and actionable information. Clustering, an unsupervised machine learning approach, is widely used to categorize students by their performance, helping educators identify at-risk students and customize appropriate interventions.
Ranjan Banerjee (Tue,) studied this question.