Analyzing course affecting the results of inferential statistics students in UiTM Seremban Campus using association rule mining / Dayang Mas Rafisha Awang Raffi, Nur Asyikin Mohd Mustafa and Nursyaza Nisa Sazali

• Main Authors: Awang Raffi, Dayang Mas Rafisha, Mohd Mustafa, Nur Asyikin, Sazali, Nursyaza Nisa
• Format: Student Project
• Language: English
• Published: 2019
• Subjects: Mathematical statistics. Probabilities; Data processing; Analysis; Analytical methods used in the solution of physical problems
• Online Access: https://ir.uitm.edu.my/id/eprint/50144/1/50144.pdf; https://ir.uitm.edu.my/id/eprint/50144/

The purpose of the study was to analyze courses taken in semester l until semester 3 that affecting the results of Inferential Statistics. Inferential Statistics was recognized as high failure rate statistics course taken by the students of Degree in Statistics. Inferential Statistics known as a most difficult course compared to the other courses since majority the students failed for the course. Lectures of University Technology Mara (UiTM) Seremban Campus faced problems with high failure rate for this course which is more than 10% students who failed for the course in every semester. Association Rule Mining was a method adopted in this study in order to achieve the objective. The second purposes of this study was to determine the interesting rules form by the independent and dependent variables .Besides , this study interested on using data mining method to extract a meaningful information and to develop relationship among the variables stored in large data sets. An Association Rule Mining was used to graduated students from matriculation and foundation and out of 427 data generated, but only 126 dataset were employed for the purpose of this study. Results showed that 239 rules have been obtained during the process of association rule mining . After removing the redundancy , only 106 rules that are most appropriate to be associated with the objectives of this research. Moreover, this study were prepared a visualization that include grouped matrix , graph-based technique and parallel coordinates plots to make it more easier in understanding the analysis. Thus , this study focused to provide an effective mechanism for lecturers in order to improve students' performance in high failure rate statistics courses.It is important for the university administration to improve in terms of the teaching and learning method in order to produce a good result for their academic performance.