Authors: Sharjil Iqbal

Abstract: In educational organizations, early evaluation prediction is a significant place of enthusiasm as it permits educators to improve performance in their courses by giving uncommon consideration at the beginning stages. This paper has proposed a model that identifies the effective features from students data for grade prediction. Paper has hide the sensitive information in the dataset to provide the privacy for the student and organization data. Feature selection is done by Modified Teacher Learning Algorithm. Work has utilized the selected feature for the training of error back propagation model for grade prediction. Experiment was done on real dataset and result shows that proposed work has improved the grade prediction accuracy.

DOI: http://doi.org/10.5281/zenodo.21279545