Analyzing the Factors that Influence Enhancing Student Performance in Oman using Data Mining

Abstract

Education field is a sign of advancement over the countries that can adopt technology to serve it. It will help to improve and enhance future achievements and be in touch with the development of technology utilizing solutions that extract student data, including their school records and other vital information about their performance, which can facilitate this process. These data are then analyzed to identify factors that affect the academic performance of the students at the school by expanding data mining techniques to enhance student academic performance. These factors are examined to develop a predictive model. Machine learning (ML) is one artificial intelligence (AI) field that can use such a model that supports educational institutions and decision-makers. A predictive method is applied using the data mining (DM) technique to take proactive action in identifying and anticipating the student's path. The data was analyzed, and the findings showed that the decision tree algorithm recorded the fastest training time for every 1000 rows. Also, the fast-scoring time for 1000 rows was in the decision tree algorithm, which was around 195 milliseconds, and the longest scoring time occurred in the random forest algorithm, which was two seconds. The top percent of classification errors reached 51% for the logistic regression algorithm and around +-1.5% of standard deviation. It took 520 mile-second for scoring time with 690 Gains for 67 m/s training time in every 1000 rows of the datasets. The findings of this study can help parents and teachers better understand the factors that influence students' academic performance and support them in assisting students with improving their academic performance.