|
|
| |
| Application of Data Mining Techniques in Students’ Performance Prediction and Analysis |
Abubakarsidiq Makame Rajab, Ramadhan Mzee Ramadhan
|
Abstract:
This study aimed to analyze academic records specifically students’ performance using data mining algorithms which can assist to explore more than one factors theoretically assumed to have an effect on students’ performance in higher education, and finds a qualitative model which quality classifies the students’ performance based totally on associated personal and social factors. The study used educational information mining to predict students' Overall GPA classification for the first year primarily based on their grades in First Semester and Second Semester results, non-public and social factors such as Living Apartment place and attendance. The study uses Data mining; Educational Data Mining (EDM), Students Performance, ID3 Algorithm, C4.5 algorithm, ID3 algorithm, CART and CHAID algorithm as tools to analyze academic records and performance of college students. The accrued data set of students' effects from the College of Health Science at the State University of Zanzibar. The study found that the discretization of the category attribute was no longer appropriate enough to capture the differences in other attributes, or, the attributes themselves was once not clear enough to capture such differences, in other words, the classes used in this study was not absolutely independent, for instance, an “Excellent” student can have the identical characteristics (attributes) as a “Very Good” student, and hence, this can confuse the classification algorithm and have large effects on its overall performance and accuracy.
This paper can be used to assist instructors with managing their class, recognize their students’ getting to know and reflect on their instructing and to assist learner reflection and provide proactive comments to learners.
|
|
|
| | |