A Comparative Study to Predict Student's Performance Using Educational Data Mining Techniques

Student's performance prediction is essential to be conducted for a university to prevent student fail. Number of student drop out is one of parameter that can be used to measure student performance and one important point that must be evaluated in Indonesia university accreditation. Data Minin...

Full description

Saved in:
Bibliographic Details
Published inIOP conference series. Materials Science and Engineering Vol. 215; no. 1; pp. 12036 - 12042
Main Authors Khasanah, Annisa Uswatun, Harwati
Format Journal Article
LanguageEnglish
Published Bristol IOP Publishing 01.06.2017
Subjects
Algorithms
Bayesian analysis
Colleges & universities
Comparative studies
Data mining
Decision trees
Education
educational data mining
Feature selection
Industrial engineering
Performance prediction
prediction
student performance
Indonesia
Online AccessGet full text

Cover

More Information
Summary:Student's performance prediction is essential to be conducted for a university to prevent student fail. Number of student drop out is one of parameter that can be used to measure student performance and one important point that must be evaluated in Indonesia university accreditation. Data Mining has been widely used to predict student's performance, and data mining that applied in this field usually called as Educational Data Mining. This study conducted Feature Selection to select high influence attributes with student performance in Department of Industrial Engineering Universitas Islam Indonesia. Then, two popular classification algorithm, Bayesian Network and Decision Tree, were implemented and compared to know the best prediction result. The outcome showed that student's attendance and GPA in the first semester were in the top rank from all Feature Selection methods, and Bayesian Network is outperforming Decision Tree since it has higher accuracy rate.
ISSN:1757-8981
1757-899X
DOI:10.1088/1757-899X/215/1/012036