A study of factors affecting learning efficiency on higher education student performance evaluation dataset using feature selection techniques

Name: Kairung Hengpraprohm.

Organization : Nakhon Pathom Rajabhat University. Faculty of Science and Technology

Name: Supoj Hengpraprohm.

Organization : Nakhon Pathom Rajabhat University. Faculty of Science and Technology

Email : supojn@yahoo.com

Name: Wannee Sudjitjoon.

Organization : Nakhon Pathom Rajabhat University. Faculty of Education

keyword: UCI machine learning dataset.

LCSH: Educational evaluation.

; Data classification.

LCSH: Data sets.

; Learning efficiency.

; Higher Education.

Abstract: This research aimed to discover the features affecting the learning efficiency on the higher education student performance evaluation dataset. The data were gathered from the student in the Faculty of Engineering and the Faculty of Education, Academic year 2019, to forecast the final learning performance of the students. Data consisted of 33 attributes and 145 records from UCI Machine Learning Dataset. Four feature selection techniques, which were Information Gain, Gain Ratio, Correlation Coefficient, and Chi-Square, were applied, along with four data classification methods: K-Nearest Neighbor, Random Forest, Artificial Neural Network, and Linear Regression. Findings demonstrated that the best feature selection techniques were Information Gain and Gain Ratio. When analyzing the relationship of feature data using Pearson’s correlation, the feature that had a positive relationship with the data class could be adapted. Further, when considering five features: CUML_GPA, EXP_GPA, READ_FREQ, COURSE ID, and KIDS: meaning when the student had a high cumulative grade point average of the last semester, high academic achievement expectation score, frequency of reading non-scientific books, and divorced or dead parents, they had satisfying learning achievement. Additionally, the attributes, which were STUDY_HRS, AGE, SALARY, IMPACT, had a negative relationship with the data class. It meant the low weekly study hour, young age, low income, and positive impact of the project or activity on the success led to satisfying learning achievement. Thus, it could be concluded that the factors affecting learning efficiency were the accumulated grade point average, achievement expectation score, frequency of reading non-scientific books, and low weekly study hours. All features could be the guideline for designing the learning management for the learner’s highest learning efficiency."""

King Mongkut's University of Technology North Bangkok. Central Library

Address: BANGKOK

Email: library@kmutnb.ac.th

Created: 2565

Modified: 2024-09-03

Issued: 2023-02-12

บทความ/Article

application/pdf

BibliograpyCitation : วารสารเทคโนโลยีสารสนเทศ. ปีที่ 18, ฉบับที่ 2 (ก.ค.-ธ.ค. 65), หน้า 34-43.

eng

©copyrights King Mongkut's University of Technology North Bangkok

ลำดับที่.	ชื่อแฟ้มข้อมูล	ขนาดแฟ้มข้อมูล	จำนวนเข้าถึง	วัน-เวลาเข้าถึงล่าสุด
1	ch36_04.pdf	460.73 KB	2	2024-07-17 19:58:30

ใช้เวลา

0.034086 วินาที