Decision support system for selecting outstanding students using simple addictive weighting (SAW) and rank order centroid (ROC) methods

Hidayatin Sholikha — 2025-09-13

Selecting outstanding students is essential in fostering appreciation and motivation within the school environment. Nevertheless, many educational institutions continue to use manual assessment methods, which are often subjective and inefficient. This research focuses on the development of a web-based decision support system designed to assist in the selection process at a in Indonesia. The system integrates the Simple Additive Weighting (SAW) technique to generate student rankings based on preference scores, while the Rank Order Centroid (ROC) method is applied to assign weight values to the evaluation criteria, including academic performance, attendance, behavior, and extracurricular involvement. Data for this study were collected through interviews, direct observation, and student records. The application was developed using PHP for the backend, MySQL for database handling, and Bootstrap for the user interface design. The system’s functionality was verified using black box testing, which confirmed that all features operated correctly. Additionally, the system was evaluated against the manual selection process conducted by the school, and the results showed an accuracy level of 80% in matching student rankings. This system proves to be a practical and structured solution for enhancing the transparency and objectivity of student achievement evaluations.

Corn sales analysis using linear regression and svm methods to improve production planning

Ahmad Hakiki Saputra — 2025-09-13

This research aimed to analyze and predict corn sales at UD Muara Kasih to improve production planning accuracy. The study used historical corn sales data collected over a specific period, covering 42 data entries from January 2021 to December 2024. The dataset included variables such as sales date, quantity sold, selling price per ton, total sales value, weather conditions, market demand (in tons), and the number of laborers. Prior to model training, the data underwent comprehensive preprocessing involving data cleaning, feature extraction, and normalization to ensure its quality and readiness for analysis. Two predictive models were applied: Linear Regression and Support Vector Machine (SVM) with a Radial Basis Function (RBF) kernel. Simulation data for 2024 and 2025 were generated based on the monthly averages derived from the historical dataset. The results showed that the Linear Regression model produced more stable predictions with a lower Root Mean Squared Error (RMSE) of 255.84 compared to the SVM model’s RMSE of 256.42. While the SVM model showed greater responsiveness to seasonal variations, the Linear Regression model was identified as the most suitable for the given dataset. The predictive models developed in this study are expected to support UD Muara Kasih in making more accurate and informed production decisions in the future.

Journal of Soft Computing Exploration

Decision support system for selecting outstanding students using simple addictive weighting (SAW) and rank order centroid (ROC) methods

Corn sales analysis using linear regression and svm methods to improve production planning