An optimal heuristic for student failure detection and diagnosis in the sathvahana educational community using WEKA

Abstract

The study offered in this paper aims to explore students characteristics and to determine unsuccessful student groups in respective subjects based on their earlier education and the impact of other factors in multiple dimensions. Predictive data mining techniques such as as classification analysis is applied in the analysis process. Datasets used in the investigation were collected from all academic years in the Sathavahana educational community contains different professional disciplines through online. The method adopted is to know the number of students failing in each subject and analyze the reasons for failure using data mining tools like WEKA. This model works effectively with large datasets. It has been tested on WEKA with different algorithms.