Prediction Model for Students’ Future Development by Artificial Intelligence Engine

Abstract

Classifying the performance of students in the examinations is one of the most difficult tasks for teachers. Though there are many traditional methods of data mining, such as the decision tree, naïve bayes and association rules which are used to classify the students’ performance, with the quick growth in technology, artificial intelligence and deep learning algorithms brought the classification and prediction methods a different approach. In this paper, we study to know how artificial intelligence algorithms can be used to identify students 'success and predict their future university and suitable program for their degree. A precise prediction is necessary to advise on the selection of a university for the student and suitable degree program. The deep learning model dataset does not only have conventional academic performance including mathematics, Chinese, English, Physics, Chemistry, Biology and History, but also non-academic performance such as service, behavior, sports, etc. Artificial neural network algorithm has few parameters which include the number of intermediate nodes and the hidden layers are adjusted and compared. The accuracy ranged from 80% to 90% with this algorithm. This model's optimum configuration reaches the maximum predictive precision that can be calculated. This study identified the factors that affect predictive model accuracy.