Predicting student performance: an application of data mining methods with an educational web-based system

Improving Academic Performance Prediction by Dealing with Class Imbalance

“…Although some papers had researched in this area such as [3], [4], and [5], they evaluated their models in terms of accuracy, which has less meaning in the case of class imbalance as we analyzed in section 2.3. We have used AUC, F-Measure, and total cost to evaluate the models on three methods proposed in section 3 and recognized that the results are reasonable.…”

“…For examples, predicting student performance, clustering similar students, and associating types of students with appropriate courses; [3], [4], and [5] used BN, DT, and other common techniques to predict the student results; another study [6] was done by us recently to predict the student performance at two real case studies: Can Tho University, Vietnam (CTU) 1 , and Asian Institute of Technology, Thailand (AIT) 2 .…”

Comparative Study: Feature Selection Methods in the Blended Learning Environment

“…Two suggested classifiers, Naïve Bayes and Neural network, were able to predict with 80% of accuracy the cases of drop out. Comparison of six classifiers was used to predict the final results of the course realized in the web system for learning [11]. The data extracted from the log files were related to the results of the solved tasks and students activity on the system (participation in communication, reading of educational material).…”

“…Some of the objectives of classifying are: anticipation of students' successfulness on the basis of data extracted from educational system web log files [11]; identification of students with low level of motivation and detection of appropriate activities in order to decrease ratio of giving up [12]; providing feedback to the students on achieved accomplishments [13]; possibility of directing and recommending learning processes in order to achieve the best results possible [14]; detecting a group of students with similar characteristics and reaction to special educational strategies [15]. There are several studies that dealt with the issue of comparison of classifiers accuracy to the data in the field of education.…”

Data Mining on Student Database to Improve Future Performance

“…In the recent years, the interest in Data Mining and Business Intelligence has boomed [1] [2]. This has led to an exponential growth in storage capacities hence leading to an increase in databases of several organizations.…”