Aina Musdholifah

This study investigates whether oversampling is a reliable solution for severe class imbalance in Arabic dialect identification. Using the Shami Corpus as a controlled testbed, we demonstrate that conventional oversampling often fails in high-dimensional sparse text spaces, but density based cluster filtering can effectively resolve this. We conduct a comparative evaluation of SMOTE, clustering-guided variants (ASTRA-SMOTE and SMOTE-RADIANT), and a cost-sensitive ClassWeight approach under an identical 5,644-dimensional feature-engineering pipeline using LightGBM and XGBoost. On the held-out test set, standard SMOTE and class weighting frequently distorted decision boundaries, yielding inconsistent gains across models. In contrast, SMOTE-RADIANT yields a statistically significant macro-F1 improvement for LightGBM (0.8539 vs. 0.8526 on the original data) with a large effect size (r = 0.511), successfully rescuing minority dialects without degrading the majority class. These findings suggest that while oversampling is not universally reliable in sparse text spaces, coupling it with density-based noise neutralization (RADIANT) provides a robust and interpretable alternative to deep learning models. This study provides methodological clarity and reproducible guidance for fair and inclusive Arabic NLP systems.

Fine Tuning Attribute Weighted Naïve Bayes (FTAWNB) is a reliable modified Naïve Bayes model. Even though it is able to provide high accuracy on ordinal data, this model is sensitive to outliers. To improve the performance of FTAWNB, this research modified the Partial Instances Reduction (PIR) technique to make the FTAWNB more adaptive to outliers. Nevertheless, in contrast to the original PIR technique, which substitutes missing values for data values deemed outliers, the PIR technique suggested in this study replaces data values deemed outliers using a Naïve Bayes weighting approach. The attribute values from the outlier data are replaced with the highest probability values for the attributes in the actual class. This PIR technique is referred to as modified PIR. The FTAWNB model with modified PIR has been evaluated using the Gaming Disorder dataset. Replacing the four attributes with the least amount of information resulted in accuracy gains of 99.74%, an increase of 1.53% over the FTAWNB model. The experimental result shows that adding the modified PIR technique to the FTAWNB model can handle the outlier in the data, proving it by increasing the performance in terms of accuracy, precision, and recall without pruning the dataset used. Doi: 10.28991/HIJ-2025-06-01-05 Full Text: PDF

Scheduling the final assignment exam is an important process that requires careful planning to ensure smooth implementation for each student. This process involves the stages of archiving final assignment submission files, determining supervisors and examiners, as well as preparing seminar and trial schedules. However, obstacles that often arise include conflicting schedules, long execution times, and low fitness values. To overcome this problem, the genetic algorithm approach is used to optimize scheduling. This algorithm can handle complex problems with a wide search space, although it has weaknesses in selecting appropriate parameters and the time required to reach the optimal solution. Genetic algorithm optimization techniques such as violated directed mutation (VDM) and tournament selection are used in this research. Previous research shows that VDM provides better results than other methods, while tournament selection improves the desired solution. It is hoped that the use of genetic algorithms with VDM and tournament selection will overcome the problem of conflicting schedules and increase the execution speed in final project exam scheduling.

Diabetes mellitus (DM) is a critical health condition caused by insulin production failure, leading to elevated blood glucose levels. DM often results in severe complications such as heart disease, stroke, and diabetic foot ulcers (DFU), which pose risks of infection and potential amputation. This study developed a machine learning model for early detection of diabetic foot ulcers, using thermogram images and the thermo dataset containing detailed foot temperature data. The multi-classifier model integrates CNNs for processing thermogram images and an ANN for tabular data analysis. Various image enhancement techniques were applied, including solarize, CLAHE, posterize, and gamma adjustment, to improve the visibility of key temperature distribution patterns. The results demonstrate that solarize consistently emerged as the most effective image enhancement method, significantly improving model performance across all evaluation metrics. Models enhanced with solarize achieved an impressive accuracy of 97.06%, alongside a perfect AUC score of 1,000. Additionally, the application of image enhancement techniques proved instrumental in reducing training and inference times, indicating computational efficiency. The integration of temperature data with enhanced thermogram images further boosted predictive accuracy while maintaining critical thermal information. This study underscores the transformative potential of image enhancement techniques, particularly solarize, in advancing the accuracy and efficiency of early detection models for diabetic foot ulcers. These findings contribute meaningfully to the development of medical imaging technologies, offering a robust framework for improving disease diagnosis and management.

Identifying Student Learning Styles Using Support Vector Machine in Felder-Silverman Model

This research focuses on predicting and analyzing rice production and yield throughout the world using ensemble learning techniques. The study applies and compares three methods: linear regression, ARIMA, and ensemble learning, to predict rice harvest yields. The results show that ensemble learning techniques significantly improve prediction performance. For instance, the ensemble model for predicting area harvested, combining Model 6 (linear regression) and Model 10 (ARIMA), achieved of coefficient of determination outperforming the individual models. Similarly, for predicting yield, the ensemble model combining Model 4 (linear regression) and Model 9 (ARIMA) achieved of coefficient of determination indicating superior prediction accuracy. For predicting production, the ensemble model combining Model 2 (linear regression) and Model 8 (ARIMA) achieved of coefficient of determination. These results demonstrate the effectiveness of ensemble learning in enhancing prediction accuracy with lower MSE and RMSE values. By analyzing various factors influencing rice yields, this research provides valuable insights for increasing rice production and yield, supporting efforts to improve the efficiency and effectiveness of rice farming, and contributing to achieving the United Nations Sustainable Development Goals (SDGs).

This research applies the fine tuning attribute weighted naïve Bayes (FTAWNB) model using ordinal data. It is known that in previous research, the FTAWNB model outperformed its competitors on the dataset used. However, the FTAWNB model has not been applied in the mental health domain that uses ordinal data. Therefore, this research used the anxiety gamers dataset to test the fine-tuning attribute weighted Naïve Bayes (FTAWNB) model. Anxiety disorders are mental health disorders that can indicate the emergence of a gaming disorder. Gamers can experience anxiety disorders classified into four classes, namely minimal, mild, moderate, and severe anxiety. Then compare the results by FTAWNB obtained with three other naïve Bayes algorithms, namely Gaussian naïve Bayes, multinomial naïve Bayes, and categorical naïve Bayes, using the same dataset. Model performance is measured based on accuracy, precision, recall, and processing time. The test results show that the FTAWNB outperforms the other three models' accuracy, precision, and recall, with an accuracy value of 99.22%. While the accuracy of Gaussian NB is 91.132%, Categorical is 91.592%, and multinomial naïve Bayes is 61.104%. However, the FTAWNB takes slightly longer than the other three models' processing time. The FTAWNB takes 0.07 seconds to build the model and 0.05 seconds to test the model on training data.