Predicting Student Performance with ANNQ3H: A Case Study in Secondary Education

Abstract

In late advancements, Educational foundations chasing and persevering best execution from the understudies. So, the Educational Institutions are endeavouring to improve the presentation of understudies in scholastics in this article an Artificial Neural Network model built using the Quick method with 3 Hidden layers (ANNQ3H) implemented in IBM SPSS modeller. It is used for predicting the performance of secondary students based on 16 parameters that are related to their regular life. For executing the ANNQ3H model data set collected from 606 secondary education students in the form questionnaire, in India. This model showed good performance with better accuracy 93.39% when compared with other Data Mining (DM) classification techniques like LR (Logistic Regression), DT (Decision Tree) algorithms, Naïve Bayes, SVM (Support Vector Machine). Applications of DM are turning into a progressively basic apparatus in comprehension and taking care of instructive and authoritative issues in training. The discoveries of this article demonstrate the viability and importance of DM techniques in-subject assessment and advanced education mining. Besides, these discoveries might be utilized to increase the estimation instruments.In addition analysis of feature importance also shown for neural network classifiers it is used for evaluate the student performance.