Sinkron : jurnal dan penelitian teknik informatika
Not a member yet
1261 research outputs found
Sort by
Lightweight Deep Learning Models for Facial Expression Recognition in Inclusive Education
No full textFacial expression recognition is an essential component in the development of artificial intelligence-based learning systems, particularly in the context of inclusive education that involves students with special needs. This study aims to evaluate the performance of several lightweight deep learning architectures in detecting facial expressions with high accuracy while maintaining computational efficiency. Facial image data were obtained from both public datasets and newly collected samples, which were preprocessed through face cropping, normalization, and data augmentation. The dataset was split into 70% training, 15% validation, and 15% testing. Four lightweight deep learning architectures: MobileNetV2, MobileNetV3 (Small and Large), and EfficientNetB0, were employed as the primary models using transfer learning and fine-tuning approaches. Evaluation was conducted using accuracy, loss, precision, recall, and F1-score metrics, complemented by visualization through confusion matrices. The results indicate that MobileNetV2 achieved the best performance with a test accuracy of 92%, precision of 93%, recall of 91%, and F1-score of 92%, while maintaining a relatively lightweight parameter size of 2.26 million. EfficientNetB0 ranked second with 83% accuracy, followed by MobileNetV3-Large (77%), whereas MobileNetV3-Small demonstrated the lowest performance (45%). Confusion matrix analysis revealed recurring misclassification patterns for certain expressions, such as Happy often misclassified as Sad, and Neutral overlapping with Angry. This study confirms that MobileNetV2 is the most optimal architecture for implementing facial expression recognition systems in inclusive education environments, as it balances high accuracy with computational efficiency. These findings provide a solid foundation for developing intelligent applications that support adaptive interaction in the learning process.
MCDM-Based Blockchain and Artificial Intelligence Integration for Earthquake Risk Recommendation System
No full textIndonesia is one of the countries with the highest earthquake vulnerability in the world because it is located at the meeting point of three major tectonic plates, namely Eurasia, Indo-Australia, and Pacific. The high risk of disaster requires a system that is capable of analyzing, predicting, and recommending earthquake-prone areas accurately, efficiently, and safely. This study aims to develop an earthquake risk recommendation system based on the integration of Artificial Intelligence (AI), Multi-Criteria Decision Making (MCDM), and Ethereum Blockchain. Earthquake data was obtained from Google Earth Engine (GEE) and geospatial data from the Geospatial Information Agency (BIG) and BMKG. The data is processed using AI algorithms for predictive analysis, then the MCDM methods of TOPSIS, and ELECTRE are applied to determine the priority of earthquake-prone areas based on a combination of seismic parameters, population density, infrastructure vulnerability, and distance to active faults. The analysis results are stored in a decentralized manner using the Ethereum Blockchain through smart contracts to ensure data integrity, security, and transparency. The research results show that the integration of AI–MCDM is capable of providing earthquake risk recommendations with high accuracy, while the application of blockchain ensures that the results cannot be manipulated. This system is expected to become a decision-making tool for disaster management agencies such as BMKG and BNPB in data-based earthquake risk mitigation
Prediction of Organic Waste Deposits in Compost Houses using LSTM and ARIMA Algorithms
No full textIndonesia faces a significant waste problem and is becoming a global challenge, mainly due to inadequate food waste management. In Kendal District, the Environmental Agency struggles to optimize waste collection and predict the volume of organic waste. To address this issue, this study explores the application of predictive technology and data analysis to improve the efficiency of waste management. Two predictive models, ARIMA and Long Short-Term Memory (LSTM), were developed and compared by collecting historical data from Kendal Organic Compost House from 2020-2024 while for train and test data using data from January 2, 2023, to December 30, 2023. The ARIMA model showed better accuracy, capturing stable trends and seasonal patterns in the time series data, with an MSE of 72,799.49. Meanwhile, the LSTM model, although capable of handling non-linear and complex patterns, performed poorly with an MSE of 54,711,498,631,770.58, indicating a failure to accommodate sharp fluctuations in the data. These findings highlight the suitability of ARIMA for data with low volatility and strong seasonality, making it more reliable for short-term predictions. The results of this study are expected to assist the Kendal District Environmental Agency in planning efficient waste management strategies, optimizing compost house operations, and improving resource allocation. Future research should focus on the integration of external variables, such as weather and population dynamics, and explore hybrid models for better prediction
Analysis of Social Assistance Donor Classification at the Muhammadiyah Medan Orphanage Using SVM
No full textThe Putra Muhammadiyah Orphanage in Medan City is a social institution that relies on donor assistance to support various social programs. The problem that occurs at the Putra Muhammadiyah Orphanage in Medan is the difficulty in identifying potential and non-potential donors who have the potential to provide sustainable social assistance contributions. This study aims to conduct a comprehensive analysis and classification of donors using the Support Vector Machine method, an effective method in machine learning to handle classification problems with SVM with high accuracy. The research data consists of donor data with several main characteristics such as the amount of donation, the frequency of donations given, and the type of assistance. The data is processed through a preprocessing stage including data normalization and data division into training and testing data. Then, the SVM model is trained to classify donors into two categories, namely Potential Donors and Non-potential Donors. Based on the data obtained from the donation bookkeeping records of the Putra Muhammadiyah Orphanage in Medan City, it can be concluded that around 55 potential donors out of 90 donors and 35 non-potential donors out of 90 donor data. From the results of the analysis and testing of the model conducted, it can be concluded that the SVM method can classify "Potential Donors" and "Non-Potential Donors" with a fairly high level of accuracy. The level of accuracy obtained reached up to 89% with a precision value of 93%, a recall value of 89% and an f1-score of 90%. With these results, this study can provide significant benefits in the management of social assistance, especially helping orphanages to determine who are potential and non-potential donors. Therefore, this study is expected to have an impact on improving the sustainability of social programs at the Putra Muhammadiyah Orphanage in Medan City
Fuzzy C-Means Algorithm for Grouping Students Based on Preferences and Academic Potential
No full textPersonalized education is increasingly becoming a necessity in the modern era to ensure that students get a learning experience that is relevant to their interests and academic potential. This study aims to group students into three main clusters, namely Science, Arts, and Business, using the Fuzzy C-Means (FCM) algorithm. The FCM algorithm was chosen because of its flexibility in handling multidimensional data and allows students to have degrees of membership in more than one cluster, reflecting the multidisciplinary nature of their preferences. The research dataset consists of data on students' interests in fields of study (Science, Arts, Business) and academic grades in related subjects. The clustering results show that: The Business cluster includes 59 students (46.9%), reflecting the dominance of interests in economics, global trend analysis, and business organization activities. Artcluster consists of 39 students (30.0%), who show a preference for visual arts, art portfolio development, and involvement in community design. Science cluster has 30 students (23.1%) with interests in biology, science experiments, and biotechnology. Evaluation using Davies Bouldin Index (DBI) yields a value of 0.78, indicating good cluster quality. In addition, manual validation from teachers shows that more than 85% of students in each cluster fit the grouping based on direct observation. This study makes a significant contribution to the development of data-driven academic recommendation systems, enabling educational institutions to design learning programs that are more adaptive, relevant, and in accordance with student needs
The Efficiency of Machine Learning Techniques in Strengthening Defenses Against DDoS Attacks, Such as Random Forest, Logistic Regression, and Neural Networks
No full textDistributed Denial of Service (DDoS) attacks are one of the most common cybersecurity concerns brought on by the quick development of digital technology. By flooding servers with too many requests, these assaults interfere with online services, highlighting the necessity of strong detection systems. Using the well-known CIC-DDoS2019 dataset, this study explores the use of machine learning algorithms—Random Forest (RF), Logistic Regression (LR), and Neural Networks (NN)—to improve DDoS assault detection. A comprehensive preprocessing procedure that comprised feature selection, normalization, and duplication removal was applied to dataset in order to ensuring optimal algorithm performance. With an accuracy of 97% on the entire test dataset and 99.13% on the training and validation datasets, RF showed exceptional performance. While NN successfully managed intricate data patterns, attaining an accuracy of roughly 94%, LR demonstrated impressive results with an accuracy of 98.65%. Because of its ensemble method, which minimizes overfitting and improves model generalization, the RF algorithm performed better than the others. This study highlights how machine learning may be used to solve practical cybersecurity issues by offering insightful information about how to optimize algorithms for real-time DDoS detection. The results improve the stability and resilience of digital infrastructures by aiding in the creation of effective intrusion detection systems. Future research can explore integrating advanced neural network architectures and hybrid methods to further improve detection rates and adaptability to evolving cyber threats
Development of RPG-Based Mathematics Educational Games with the Waterfall Method on Fraction Material for Elementary School Students
No full textMathematics is one of the subjects that plays an important role in everyday life and in developing logical thinking and problem-solving skills. One of the topics that often poses a challenge for elementary school students is fraction numbers. Fractions are often considered an abstract and difficult concept to understand because they involve numerical representations that differ from whole numbers. This difficulty frequently leads to a lack of interest in learning mathematics, ultimately affecting students' academic performance. The data collection stages applied include interviews, observations, and the distribution of questionnaires. The development of this learning media follows the Waterfall model, which aims to design improvements to the existing system. The results of the User Acceptance Test reveal that this game received a user perception score of 90.5%, categorizing it as "very good," indicating that students find it both enjoyable and effective as a learning tool. This suggests that the game is not only engaging but also effective in helping students understand fraction concepts in a more interactive and enjoyable way. With the presence of story elements, challenges, and engaging game mechanics, students can learn in a more immersive manner compared to conventional methods. Therefore, this game is suitable for use as a mathematics learning tool, particularly in understanding fraction operations such as addition, subtraction, multiplication, and division
Collective Intelligence for Cybersecurity: Federated Learning under Non-IID Conditions for Intrusion Detection
No full textCyber threats are becoming increasingly complex in cyberspace, which highlights the necessity for strong Intrusion Detection Systems (IDS). However, traditional centralized IDS methods have large problems with data privacy and scalability. Federated Learning (FL) is an intriguing new idea that lets multiple clients train a model together without sharing data directly, which keeps privacy intact. The proposed federated intrusion detection model develops and assesses FL models for detecting network intrusions, focusing on the important issue of non-independent and non-identically distributed (non-IID) data among clients. This work implements and compares two widely recognized FL algorithms, Federated Averaging (FedAvg) and Federated Proximal (FedProx), using a 1D Convolutional Neural Network (CNN) architecture specifically designed for tabular network traffic data. The authors utilize a Dirichlet distribution (α=0.1) to distribute the data among 10, 20, and 30 clients, thereby simulating non-IID conditions in the experiment. The authors thoroughly compare the performance of algorithms using two benchmark datasets: NSL-KDD and NF-Bot-Net-V2. The comparison reveals that while both FedAvg and FedProx achieve high detection rates on NSL-KDD, FedProx is more capable of maintaining stability and converging on the more complex NF-Bot-Net-V2 dataset, achieving an accuracy of 0.9953. The results highlight that FedProx is a more appropriate algorithm for implementing robust and privacy-preserving federated intrusion detection systems in statistically heterogeneous network environments found in the real world
Mobile Banking Service Quality and User Loyalty Using MSQUAL: A Systematic Literature Review
No full textDigital transformation has made mobile banking a core service in the banking industry, emphasizing service quality as a critical factor for user satisfaction and loyalty. This study presents a systematic literature review (SLR) of mobile banking research from 2021 to 2025, guided by PRISMA and structured using the PICOC framework (Population, Intervention, Comparison, Outcome, Context) to systematically select and evaluate relevant studies. The MS-QUAL model, comprising nine dimensions: efficiency, system availability, responsiveness, privacy, content, contact, billing, fulfillment, and compensation, was used as the evaluation framework. Out of 924 initially identified articles, 20 met the inclusion criteria for in-depth analysis. Findings show that efficiency, system availability, privacy, responsiveness, content, and fulfillment consistently drive user satisfaction, while compensation, contact, and billing have limited influence. Satisfaction serves as the primary mediator connecting service quality to loyalty, indicating that improvements in MS-QUAL dimensions must translate into positive user experiences to foster long-term loyalty. The study further highlights challenges in maintaining security standards, adapting traditional dimensions to evolving user expectations, and ensuring consistent service quality. Opportunities lie in leveraging technologies such as AI, blockchain, and big data to create personalized, secure, and interactive experiences, enhancing both functional and emotional engagement. Overall, MS-QUAL remains a relevant and flexible framework for evaluating mobile banking service quality when aligned with contemporary technological advances and user-centered strategies
Hybrid GA–MILP Model for Community Building Retrofit Planning Towards Carbon Neutrality
No full textRetrofitting community buildings is a key pathway toward carbon neutrality, yet most existing retrofit planning models lack adaptability to the diverse urban contexts of the Global South, where building typologies are heterogeneous and resources limited. Addressing this gap requires approaches that are both computationally efficient and context-sensitive. This study introduces a hybrid optimization framework that integrates Genetic Algorithm (GA) and Mixed-Integer Linear Programming (MILP) to tackle the multidimensional multiple-choice knapsack problem inherent in retrofit planning. The GA explores high-level system configurations, while MILP ensures precise component-level selection under budget and technical constraints. Compared to conventional single-method approaches, the hybrid GA–MILP achieves near-optimal emission reduction with reduced computation time and greater feasibility, offering a balanced trade-off between performance and scalability. Importantly, the framework demonstrates that medium-cost retrofit strategies provide the most cost-effective path to scalable carbon savings, making it highly relevant for resource-constrained urban environments. By situating retrofit planning within the realities of the Global South, this study advances methodological innovation and provides a robust decision-support tool aligned with sustainable development goals for inclusive and low-carbon urban futures