Abstract
There is an increasing demand for predicting the students' performance in undergraduate level to help the decision makers to plan the admission of students to the various specializations. Whereas Prince Sattam University (PSU) and Sana'a University (SU) have large databases of students' information, such databases are not used effectively for decision-making. One of such difficulties stems from the weakness of some statistical methods in analyzing large volumes of data, resulting in unreliable outputs and results. In order to overcome these difficulties, alternative methods are constantly developed by researchers. As data mining simulates the role of a data analyst in exploring the relationships between data as authenticated in the actual data records, one of the most popular data mining techniques is the predictive model. Curve estimation model (CEM) stands by far as the most important predictive modeling technique. The CEM is used to build a highly efficient regression model to boost the accuracy of the results which can then be generalized. This paper mainly aims to build an accurate regression model to predict the extent to which the mean of admission criteria (MAC) contributes to the academic achievement for the two universities' students. The proposed model will be of help to university decision-makers in predicting students with more likelihood of poor academic attainment and helping such students improve the level of their educational attainment or to redirect them to specializations suiting their abilities. In terms of methodology, the investigation started by collecting the datasets, followed by data preprocessing stage, model establishment stage, and pattern evaluation stage. The results confirmed that the cubic model was the best model for both the universities. Further, the decision-makers at both universities are advised to use the cubic model when forecasting the students' academic achievement based on the mean of admission criteria. Also, results showed that it could identify at-risk students in a timely manner and improve the overall efficiency and effectiveness of education management. Further, the decision-makers in all the universities are recommended to utilize the data mining techniques based on the CEM, being the potential mean to analyze very large size records in quite short time.