Advancement Through Research
One of the most important challenges in modern society is related to university students’ involvement in different paths which may lead to their educational and occupational failures. To have a good career in the futures, students might choose to study majors and courses for which they do not have sufficient abilities. Consequently, they may face educational failures. Thus, classifying students’ levels and conditions based on certain features can help them significantly improve their educational quality. Indeed, although different methods have been proposed for classifying students’ information, some of them are not sufficiently accurate and precise. In this paper, a hybrid model based on naïve Bayes, honey bee colony, rough sets and K-means clustering has been proposed. The first target of the proposed hybrid method was high accuracy and accuracy in predicting the class of a new instance; the second target of this hybrid method was to extract student behaviors and detect effective factors regarding students’ educational levels. In general, it can be argued that the primary purpose of the proposed hybrid method was to enhance and improve the estimation capability of the model so that it can allocate each data to appropriate category. The evaluation of the proposed hybrid method on standard datasets indicated the optimization and improvement of the accuracy of the proposed predictive method and some other criteria.
Naïve Bayes; Educational Quality; Classification; Honey Bee Colony
