Analyzing student academic performance across streams at SMK Seri Pagi using K-Means Clustering

The academic performance of students is a critical concern for schools, yet many institutions such as SMK Seri Pagi continue to rely heavily on manual spreadsheet analysis, which is time-consuming and lacks visual clarity. This project aims to improve the analysis of student academic performance across streams by implementing a descriptive data analytics solution using K-Means clustering and interactive visualization. Following the CRISP-DM methodology, data preprocessing was conducted using Microsoft Excel and Google Spreadsheet, while RapidMiner was employed for K-Means clustering analysis. The clustering model utilized selected attributes including Gender, Stream, marks for Bahasa Melayu, Bahasa Inggeris, Matematik, Sejarah and the Average Grade Point (GP). The optimal number of clusters was determined using the Davies- Bouldin Index. The final Power BI dashboard features multiple interactive pages presenting overall statistics, clustering insights, and subject-wise analysis. Expert evaluation rated the dashboard positively for clarity and usefulness. The dashboard significantly improved the school's ability to visualize and interpret student performance patterns, facilitating more efficient and informed decision-making by educators. Challenges related to data preprocessing and integration were addressed through iterative refinement and the use of complementary tools, ensuring data quality and reliability. This project offers a data-driven approach that supports teachers in interpreting student performance patterns and making informed decisions. Future work may involve integrating real-time data updates, expanding clustering attributes, and enhancing dashboard interactivity for broader educational applications.