Kinetik: Game Technology, Information System, Computer Network, Computing, Electronics, and Control

Integrating Adaptive Sampling with Ensembles Model for Software Defect Prediction

Muhammad Yusuf, Arinal Haq, Siti Rochimah — Thu, 08 May 2025 00:00:00 +0000

Handling class imbalance is a challenge in software defect prediction. Imbalanced datasets can cause bias in machine learning models, hindering their ability to detect defects. This paper proposes an integration of Adaptive Synthetic Sampling (ADASYN) and ensemble learning methods to improve prediction accuracy. ADASYN enhances the handling of imbalanced data by generating synthetic samples for hard-to-classify instances, while the ensemble stacking technique leverages the strengths of multiple models to reduce bias and variance. The machine learning model used in this study is K-Nearest Neighbors (KNN), Decision Tree (DT), and Random Forest (RF). The results demonstrate that ADASYN combined with ensemble stacking outperforms the traditional SMOTE technique in most cases. For instance, in the Ant-1.7 dataset, ADASYN achieved a stacking accuracy of 90.60% compared to 89.32% with SMOTE. Similarly, in the Camel-1.6 dataset, ADASYN achieved 91.56%, slightly exceeding SMOTE’s 91.32%. However, SMOTE performed better in simpler models like Decision Tree for certain datasets, highlighting the importance of choosing the appropriate resampling method. Across all datasets, ensemble stacking consistently provided the highest accuracy, benefiting from ADASYN's adaptive resampling strategy. These results underscore the importance of combining advanced sampling methods with ensemble learning techniques to address class imbalance effectively. This approach improves prediction accuracy and provides a practical framework for reliable software defect prediction in real-world scenarios. Future work will explore hybrid techniques and broader evaluations across diverse datasets and classifiers.

Clustering of High School Quality Using Fuzzy C-Means in the Special Region of Yogyakarta Province

Lilin Rofiqatul Ilmi, Haryanto, Maria Sumunaringtyas — Thu, 08 May 2025 00:00:00 +0000

This research aims to reveal the results of clustering high school quality using fuzzy c-means in the Special Region of Yogyakarta Province. This research is quantitative and descriptive. Data collection was conducted through documentation. The research data are secondary data from the 2023 high school education report card. The sample consisted of 51 schools, which were determined using the proportional stratified random sampling. Data analysis was performed using the quantitative descriptive method and fuzzy c-means. The results of the study are clustering on the main indicator data producing three clusters: cluster 1 consists of 11 private schools accredited A and B, cluster 2 consists of 22 public and private schools accredited A, and cluster 3 consists of 18 schools accredited A, B, and C. Cluster 2 excels with the overall best performance, cluster 1 has moderate performance with several areas needing improvement, such as instructional leadership, the use of information technology for budget management, and inclusiveness, and cluster 3 shows the lowest performance, requiring significant attention and improvement in almost all aspects, especially literacy, numeracy, instructional leadership, and the use of information technology for budget management. Cluster 3, which had the lowest performance, showed an urgent need for improvement in almost all aspects.

Optimized BiLSTM-Dense Model for Ultra-Short-Term PV Power Forecasting

Christianto Tjahyadi, Nana Sutarna, Prihatin Oktivasari — Thu, 08 May 2025 00:00:00 +0000

The growing integration of photovoltaic (PV) systems into power grids poses challenges due to the inherent variability in PV output, particularly during rapid weather changes. While existing forecasting methods often struggle to capture these fluctuations, accurate ultra-short-term PV power prediction is critical for grid stability. The study aims to develop an optimized BiLSTM-Dense model that enhances forecasting accuracy by incorporating an additional dense layer. The model is designed to improve forecasting performance over a 30-second horizon. It utilizes a dataset of solar irradiance, PV output power, surface temperature, ambient temperature, humidity, and wind speed, collected in late 2023. Data preprocessing involved normalization and smoothing techniques to enhance robustness. Hyperparameter optimization was performed using grid search. Evaluation results demonstrate the superiority of the proposed model, achieving an MAE of 0.00271 and an RMSE of 0.00806 when paired with the Adam optimizer and Swish activation function. Compared to standard BiLSTM, the BiLSTM-Dense achieved MAE and RMSE improvements of 0.52% and 2.19%, respectively. It also outperformed the LSTM model with reductions of 4.00% in MAE and 2.65% in RMSE, and significantly surpassed ARIMA, reducing MAE by 98.87% and RMSE by 97.21%. These findings highlight the model’s ability to capture complex, non-linear dependencies in PV output data, outperforming conventional approaches like ARIMA, which rely on linear assumptions, and simpler architectures like LSTM, which lack bidirectional context integration.

Performance Evaluation of Outgoing Interface Selection Method on Fortigate SD-WAN for Network Optimization

Mufti Kholil Romadhoni, Larynt Sawfa Kenanga, Denar Regata Akbi, Diah Risqiwati — Thu, 08 May 2025 00:00:00 +0000

Reliable and high-performance network services are essential to facilitate communication between parent companies and subsidiaries as well as among the subsidiaries themselves. Challenges arise in managing and optimizing outgoing interface selection in an effective and reliable Software-Defined Wide Area Network (SD-WAN) environment. This research evaluates four outgoing interface selection methods, namely Manual, Best Quality, Lowest Cost, and Maximize Bandwidth (SLA), using a tree-based network topology simulated in GNS3 with FortiGate devices as part of the simulation. The results show that under simulated disturbances, such as limiting a single connection line to 10 kbps, the Manual, Best Quality, and Lowest Cost methods perform worse than the Maximize Bandwidth method. In contrast, the Maximize Bandwidth method outperformed the others, achieving only 0% packet loss, 22.275 ms one-way delay, and 7.03 ms jitter, while maintaining the ITU-T G.1010 standard at the preferred level. These findings highlight the reliability and effectiveness of the Maximize Bandwidth method in ensuring consistent data transmission even under fault conditions, while providing practical guidance for network engineers in configuring SD-WAN for uninterrupted high-quality network services in complex business environments.

Improvement Improvement of AC Bus Voltage Stability with Current Control Inverter

Bayu Rahmad Nugroho, Reza Maulidin, Adhi Kusmantoro — Thu, 08 May 2025 00:00:00 +0000

This research focuses on the development and analysis of a current control method for inverters, which demonstrates superior performance compared to the more conventional voltage control method. Current control in inverters offers several significant advantages, including faster dynamic response, constant switching frequency, and the ability to effectively reduce harmonic distortion, which is often a challenge in modern power systems. Additionally, this method is capable of maintaining system stability even when it had complex load variations and fluctuating operating conditions. In this study, we implement a fuzzy logic approach to simulate current control in an inverter integrated with a photovoltaic (PV) renewable energy system. The simulation results indicate that the proposed current control method not only enhances overall energy efficiency, but also extends the operating range of the inverter, allowing the system to operate optimally under various load conditions.

Design and simulation of battery charging system with constant temperature – constant voltage method

Thu, 08 May 2025 00:00:00 +0000

Batteries are essential to many contemporary applications, including electric cars and portable electronics. Overheating and charging time efficiency are the two biggest issues with battery charging. Overheating presents safety hazards and hastens battery deterioration. Due to their inability to regulate temperature, conventional charging techniques like Constant Current - Constant Voltage (CC-CV) result in excessive temperature rises during battery charging, which shortens battery life. A novel approach that helps lessen excessive temperature rises is the Constant Temperature - Constant Voltage (CT-CV) method, according to researchers. In order to avoid excessive temperature increases during the initial charging, the CT technique initially regulates the applied temperature. Second, to guarantee full capacity without causing damage to the battery, the CV technique is used to maintain a steady voltage. A fuzzy logic controller (FLC) control system is used to regulate the temperature and current at the DC-DC converter's output. The FLC control system's goal is to control the duty cycle such that the buck converter's output is 65V 11.5A. The simulation results show that the CT-CV method can reduce the increase in temperature in the battery with an average temperature during the battery charging process of 23.57 ° C with fuzzy control and 23.71 ° C with PI control. In addition, by comparing two control systems with the CT-CV method, namely PI and fuzzy, it was found that the fuzzy method was able to accelerate battery charging by 4.16% compared to the PI control.

Development of Lung Cancer Risk Screening Tool with Causal Discovery Model Evaluation Approach

Thu, 08 May 2025 00:00:00 +0000

Causal graph discovery approaches in healthcare for detecting high-risk diseases have been more widely applied in the last decade. The main challenge in causal graph discovery in healthcare data is the complexity of big data, which requires appropriate algorithms to reveal causal relationships between variables. This study focuses on evaluating the performance of seven causal discovery models—Peter-Clark (PC), Greedy Equivalent Search (GES), Direct LiNGAM, Directed Acyclic Graph-Graph Neural Network (DAG-GNN), Greedy Sparsest Permutation (GraSP), and Recursive Causal Discovery (RCD)—on opensource healthcare datasets. The model performance was evaluated using the Structural Intervention Distance (SID), Structural Hamming Distance (SHD), Matthews Correlation Coefficient (MCC), and Fobernius Norm (FN) metrics. The evaluation results conclusively show that the GES model performs best on low-complexity datasets. Meanwhile, the DAG-GNN model offers consistent performance on high-complexity data with MCC values ranging from 0.77 to 0.88. The application of the GES model for lung cancer risk screening, based on user question responses, demonstrated effectiveness by measuring MCC, SID, and SHD scores between the reference adjacency metrics and the resulting screening metrics.

Bamboo Diameter Detection System Based on Image Processing as a Pre-Assessment for an Automated Bamboo Splitting Technology

Thu, 08 May 2025 00:00:00 +0000

Bamboo is recognized for its eco-friendly attributes and rapid growth, serves as a promising sustainable alternative to wood. However, the high production cost of laminated bamboo remains a major challenge due to labor-intensive processes, particularly manual splitting, which affects efficiency and labor costs. To overcome this issue, this study presents an automated bamboo diameter measurement system that leverages Canny Edge Detection and Hough Transform to ensure precise and uniform slat dimensions. A dataset of 100 bamboo images with diameters ranging from 11 - 13 cm was utilized for training and testing. The system achieved a high accuracy, with a coefficient of determination (R²) of 0.973, demonstrating strong predictive reliability. Furthermore, Bayesian Optimization was applied to fine-tune parameters, resulting in an optimized configuration for both Canny Edge Detection and Hough Transform. The proposed system reduces dependence on manual labor, thereby lowering production costs and improving overall manufacturing efficiency. Automation in the bamboo splitting process ensures consistent and precise slat dimensions, supporting scalability and enhancing the economic feasibility of laminated bamboo production. The findings of this study provide a practical and sustainable solution to optimize production, making laminated bamboo a more viable and competitive material in the industry.

Sentiment Analysis on Social Media Uusing CNN-RNN Hybrid: A Case Study of Indonesian Presidential Candidate

Thu, 08 May 2025 00:00:00 +0000

Research on sentiment analysis for Presidential Candidate 01 on social media cannot be ignored because there is no in-depth understanding of public perceptions and opinions circulating online. The CNN model is quite commonly used for sentiment analysis; however, this model still has quite low accuracy so modifications need to be made. This research aims to increase the accuracy of sentiment analysis through the application of a modified Convolutional Neural Network (CNN) method. The research process includes collecting tweet data related to Presidential Candidate 01 using crawling techniques, data preprocessing, sentiment labeling, data balancing, as well as dividing the dataset into training, validation and test data. The CNN model is modified with additional layers to improve the performance. The model is evaluated by measuring its accuracy, precision, recall, and F1 Score. The research results show that the modified CNN-RNN Hybrid model with the Upsampling method achieves an accuracy of 94% and F1 Score of 0.95, while the CNN-RNN Hybrid model has an accuracy of 86% and F1 Score of 0.82, the CNN Model has an accuracy of 90% and F1 Score of 0.88, and the RNN model has an accuracy of 88% and F1 Score of 0.84, which are higher compared to the Naïve Bayes and LSTM methods used in the previous research. Modifying the CNN method can significantly increase the accuracy of sentiment analysis for Presidential Candidate 01, so that it can become a more effective tool for understanding public perceptions and improving political campaign strategies.

Optimizing Connected Vehicle Routing Protocol for Smart Transportation Systems

Anggiet Harjo Baskoro Bonari, Ketut Bayu Yogha Bintoro — Thu, 08 May 2025 00:00:00 +0000

The significant growth in integrating connected vehicles into intelligent transportation networks has underscored the importance of Vehicle-to-Vehicle (V2V) communication in optimizing route efficiency, reducing traffic congestion, and enhancing road safety. However, routing protocols such as AODV face substantial challenges in dynamic automotive environments characterized by high mobility and rapid topology changes, leading to issues like packet loss, delays, and network congestion. Reactive protocols like AODV often suffer from route discovery delays, while proactive protocols like DSDV, although reducing latency, increase bandwidth consumption, making them less effective in highly dynamic contexts. This study introduces the Learning Automata Ad Hoc On-Demand (LA-AODV) routing protocol, designed to improve relay node selection and V2V communication efficiency. The proposed method leverages real-time vehicle data to predict and select optimal relay nodes under dynamic traffic conditions, thereby enhancing packet delivery ratio, throughput, and reducing latency and routing overhead. The results demonstrate that LA-AODV significantly outperforms AODV and DSDV across various traffic scenarios, with an increase in packet delivery ratio up to 4% in high traffic conditions, throughput reaching 125 units, and a reduction in end-to-end delay within the range of 2E+10 to 6E+14. These improvements highlight LA-AODV's superior efficiency in handling packet loss and latency, making it a suitable protocol for data-intensive and safety-critical applications that demand reliable and efficient data transmission. This study contributes by developing the LA-AODV protocol, which significantly enhances V2V communication performance in dynamic traffic scenarios and provides a robust simulation model replicating real-world conditions, potentially reducing traffic accidents.

Voltage Control of Inverter for Microgrid Power Enhancement Using PID

Reza Maulidin, Bayu Rahmad Nugroho, Adhi Kusmantoro — Thu, 08 May 2025 00:00:00 +0000

An inverter is a device that converts direct current (DC) into alternating current (AC), which is crucial in various applications, including solar power systems, uninterruptible power supplies (UPS), and electric motor control. Accurate and stable voltage control of the inverter is essential to ensure the performance and reliability of the system. The Proportional-Integral-Derivative (PID) control method is one of the most commonly used control techniques due to its simplicity and effectiveness across different control systems.

This study focuses on the implementation of inverter voltage control using a PID controller. The PID controller is designed to regulate the inverter's output voltage, ensuring stability even in the presence of disturbances or load variations. In this research, the mathematical model of the inverter and the PID control system is developed and simulated using MATLAB/Simulink software.

The simulation results demonstrate that the PID controller effectively maintains the inverter's output voltage, providing a rapid transient response with minimal overshoot. The application of the PID controller to the inverter also shows improvements in system stability and a reduction in steady-state error. Furthermore, precise tuning of the PID parameters is a key factor in achieving optimal control performance.

This research makes a significant contribution to the field of inverter control by demonstrating the effectiveness of the PID controller in regulating the inverter's output voltage. The practical implementation of PID controllers on inverters is expected to enhance the efficiency and reliability of power systems that utilize inverters.

Optimizing Autonomous Navigation: Advances in LiDAR-based Object Recognition with Modified Voxel-RCNN

Thu, 08 May 2025 00:00:00 +0000

This study aimed to enhance the object recognition capabilities of autonomous vehicles in constrained and dynamic environments. By integrating Light Detection and Ranging (LiDAR) technology with a modified Voxel-RCNN framework, the system detected and classified six object classes: human, wall, car, cyclist, tree, and cart. This integration improved the safety and reliability of autonomous navigation. The methodology included the preparation of a point cloud dataset, conversion into the KITTI format for compatibility with the Voxel-RCNN pipeline, and comprehensive model training. The framework was evaluated using metrics such as precision, recall, F1-score, and mean average precision (mAP). Modifications to the Voxel-RCNN framework were introduced to improve classification accuracy, addressing challenges encountered in complex navigation scenarios. Experimental results demonstrated the robustness of the proposed modifications. Modification 2 consistently outperformed the baseline, with 3D detection scores for the car class in hard scenarios increasing from 4.39 to 10.31. Modification 3 achieved the lowest training loss of 1.68 after 600 epochs, indicating significant improvements in model optimization. However, variability in the real-world performance of Modification 3 highlighted the need for balancing optimized training with practical applicability. Overall, the study found that the training loss decreased up to 29.1% and achieved substantial improvements in detection accuracy under challenging conditions. These findings underscored the potential of the proposed system to advance the safety and intelligence of autonomous vehicles, providing a solid foundation for future research in autonomous navigation and object recognition.