<![CDATA[College is the highest educational institution that is important in preparing students to face global challenges. However, not a few students experience academic failure in completing their lectures (drop out). The high dropout rate can be minimized by making the right decisions to prevent students from dropping out. To help make these decisions, a technology such as Educational Data Mining (EDM) is needed to better understand student data patterns. One method in EDM that can be used is data classification with the concept of K-Nearest Neighbor (K-NN). In this research, K-NN works by analyzing the closeness of student data to predict the likelihood of a student graduating on time or at risk of dropout. The data used covers aspects of student life, such as academic, social, and economic aspects. Based on the results of the analysis, it can be concluded that the use of K-NN has good performance and is quite accurate but can be improved by adding resampling techniques to overcome class imbalances that may occur from the influence of large amounts of data. However, the K-NN approach can help universities design policies because it is simple to implement, making it easy to apply in the academic field. Keywords: Educational Data Mining (EDM), K-Nearest Neighbors (KNN), Drop Out, College.]]>
Copyrights © 2025
