Higher Education Management and Student Achievement Assessment Method Based on Clustering Algorithm

Area of Science:

• Educational Data Mining
• Higher Education Management
• Instructional Management

Background:

• Traditional student performance evaluation methods have limitations in reflecting comprehensive student data.
• Data mining offers advanced analytical capabilities for educational data.
• Effective instructional management relies on timely and accurate student performance analysis.

Purpose of the Study:

• To propose an improved K-means clustering algorithm for student performance evaluation.
• To address challenges in evaluating student performance due to varying course difficulties.
• To enhance the effectiveness of higher education management through data-driven insights.

Main Methods:

• Utilized the K-means clustering algorithm for student performance data analysis.
• Developed and investigated an improved K-means algorithm incorporating student information.
• Compared the proposed algorithm's performance against traditional K-means and fast global mean clustering.

Main Results:

• The improved K-means algorithm demonstrated significantly faster clustering times (0.04) compared to other methods.
• All tested clustering algorithms successfully grouped data, including noisy points.
• The proposed algorithm shows practical advantages in processing student performance data.

Conclusions:

• Clustering algorithms, particularly the improved K-means, offer a valuable mechanism for higher education management and student performance evaluation.
• This approach provides insights into student learning patterns for better instructional guidance.
• The findings support data-driven decision-making in educational administration.