Transcript PowerPoint 프레젠테이션

Predicting Undergraduate Students
Success using Logistic Regression
technique
Apichai Trangansri, Luddawan Meeanan, Settachai Chaisanit and
Anongnart Srivihok
Faculty of Information Technology, Sripatum University Chonburi Campus,
Thailand
Outline
Introduction
The Objective
Data set
Literature Review
Methodology
Results
Conclusion
INTRODUCTION
• The concept of Education
• Predictive model
• The factors for Predictive
INTRODUCTION
• The aim of this study was to predictive model for
undergraduate students’ success. It provides a
manageable structure for the administration of
admission of new students and learning
management in the institution. Moreover, the
predictive model creating knowledge and
strategies to improve teaching and learning
management in Thailand.
The Objective
to explore the predicting undergraduate
student success using Logistic Regression
technique
Population and Dataset
• Population of this studied was comprised of :
– The populations are undergraduate students at Sripatum
University Chonburi Campus, Thailand.
• Dataset:
– 3, 719 dataset
Literature Review
• Forecasting Techniques
• Forecasting techniques are typically broken
into the categories of time series, regression,
and subjective techniques.
• Logistic Regression
•
Logistic Regression analysis is used for prediction
of the probability of occurrence of an event by
fitting data to a logit function logistic curve.
Logistic Regression
Methodology
• The methodology of this study comprise of the
A logistic regression model was built using
data from Sripatum University Chonburi
Campus, Thailand. The applicants from the
2001-2011 acadamic years.
Methodology
• Datasets were obtained from the Faculty of
Business
Administration,
Faculty
of
Accounting
and
Faculty
of
Information
Technology. The sample group of this study
was 3, 719 dataset.
This research has been
divided into 3 classes and 9 variables.
Methodology
RESULTS
100
90
80
70
60
50
40
30
20
10
0
Faculty of Business
Administration
Faculty of Accounting
Faculty of Information
Technology
Prediction model
RESULTS
• The prediction model divided by faculty show
ed that:
• Faculty of Business Administration: GPA
increased by one unit, the students has oppor
tunity to graduated 81.5 percent.
• Faculty of Accounting: GPA increased by one
unit,
the
students
has
opportunity
to
graduated 64.3 percent.
• Faculty of Information Technology: GPA
increased by one unit, the students has
opportunity to graduated 80.8 percent.
Conclusion
• This research applied of data mining technique for generate Predictive
Modeling of undergraduate students success. The research results
supported the idea that the ways in which student success can be
predicted in conventional and education.
• Therefore, A logistic regression model was built using data on the
applicants from the 2001-2011 acadamic years at Sripatum University
Chonburi Campus, Thailand. The result showed that the relationship of
variables showed that the data field: major, GPA, age and gender are
variables that affect the student succes in significant at 0.05.
• However, the benefits of predictive model for undergraduate students
success. It provides a manageable structure for the administration of
admission of new students and learning management in the institution.
Moreover, the predictive model creating knowledge and sstrategies to
improve teaching and learning management in Thailand.
Thank you