Abstract

• The prediction of Basic Education High School students’ dropout has been an important field for educational institutions. Recently, Educational Data Mining (EDM) has gained attention among educational researchers and information technology researchers. Developing a strategic plan helps every student to attend school with positive outcomes. Data mining technique called cluster can predict why students drop out. The proposed system of this study is to analyse the performance of data mining techniques and to predict students' dropout using the K-Means clustering algorithm. A pre-processing step for the student; parent; teacher survey data led to a higher level of accuracy due to data cleaning and data reduction using Principal Component Analysis. The proposed system processed the survey data of the Basic Education High School students in rural Pathein, Ayeyarwady Region.