Self Organizing Map (SOM) based Modelling Technique for Student Academic Performance Prediction

Abstract

Over the years, student academic performance mapping is considered an important issue for academic institutions and designing such system is very complicated. However, the student performances rely on various factors such as attendance, marks, family background, curriculum activities, social behavior etc. and mapping of all these attributes is very complicated. In the past, various data mining software and techniques have been proposed to classify student data set. These software�s and techniques have been failed to classify student dataset correctly. Now advances of Artificial Intelligence (AI) and data mining techniques made it possible to classify student data set and draw useful patterns efficiently. In this study, real data set of Government Girls College (GGC) vidisha of 250 students is considered. The main concern of this study is to apply SOM clustering approach to classify student dataset. Finally, experimental results demonstrated that 4 clusters have been formed based on category like very good, good, average, and poor.