ejournal.nusamandiri.ac.id (STMIK Nusa Mandiri)
Not a member yet
1504 research outputs found
Sort by
WASTEWISE: AI-POWERED WASTE EDUCATION FOR ELEMENTARY STUDENTS USING YOLOV8 AND ESP32-CAM
Full text linkThe growing volume of global waste poses a significant challenge for effective waste management, particularly in developing countries where awareness and practices around waste sorting remain limited. This study aims to enhance elementary school students' understanding and efficiency in sorting organic and inorganic waste using an interactive, AI-powered educational tool. The proposed system, WasteWise, integrates YOLOv8 for real-time object detection and ESP32-CAM for capturing waste images. A pre-test and post-test experimental design was conducted to assess students’ performance before and after using the system. The results showed a notable improvement in sorting accuracy, increasing from 60% with manual sorting to 90% using the WasteWise system, alongside reduced sorting time. These findings highlight the system's potential not only as an automated waste classification tool but also as a cost-effective and engaging platform for promoting environmental awareness and digital literacy among young learners
IDENTIFICATION OF FOOD DIVERSIFICATION ON JAVA ISLAND USING ARCGIS
Full text linkIndonesia is addressing the challenges of food security and consumer preference also known as Food diversification. The research aims to analyze the potential of various local food sources as alternatives to rice, which is the dominant staple food in Indonesia, with a particular focus on geographic implications. Although local carbohydrate sources like corn, potatoes, and tubers are available, their adoption is limited and understudied in relation to geographic distribution and consumer behavior. This study integrates survey data and GIS-based spatial analysis to evaluate local food diversification potential. Findings show that while 100% of respondents consume rice, 48.7% have tried alternatives, with limited availability (41.03%) and higher costs (17.95%) as key barriers. With 94.7% expressing willingness to adopt new staples, the results suggest GIS-based decision support systems can guide effective, region-specific food policy interventions
ISOLATION FOREST PARAMETER TUNING FOR MOBILE APP ANOMALY DETECTION BASED ON PERMISSION REQUESTS
Full text linkEnsuring mobile app security needs the capability to detect apps that request excessive or inappropriate permissions. This research proposes an anomaly detection approach using Isolation Forest, enhanced through hyperparameter tuning, to identify suspect apps based on permission request patterns. The dataset is processed into binary features, followed by exploratory data analysis (EDA) to examine the distribution and highlight sensitive permissions. The Isolation Forest model is then optimized by tuning parameters such as contamination level, number of estimators, and sample size. The fine-tuned model achieved a more accurate separation between normal and anomaly applications, detecting 10 anomalies out of 200 applications, with anomaly applications averaging 125.10 permits compared to 42.76 in normal applications. These anomalies often requested permissions related to network, storage, contacts and microphone, indicating potential privacy risks. The results show that parameter tuning improves the detection performance of Isolation Forest, providing a practical solution for mobile security monitoring. After tuning, the number of false positives decreased by 50%, and the model successfully reduced detected anomalies from 20 to 10, increasing the precision of anomaly detection from 70% to 90%. Future work could include improving feature selection and integration into real-time detection systems.
SENTIMENT ANALYSIS OF PUBLIC OPINION ON TRANSPORTATION SERVICES IN INDONESIA USING MACHINE LEARNING
Full text linkThis study analyzes public sentiment towards transportation services in Indonesia through social media using Naïve Bayes and Support Vector Machine (SVM) algorithms. Data was collected from Twitter using an API with transportation-related keywords over a three-month period. The analysis results indicate that 93.5% of the opinions are neutral, 3.5% are positive, and 3% are negative. The dominance of neutral sentiment suggests potential dataset imbalance or user hesitation in expressing strong opinions. SVM achieved a higher accuracy (100%) compared to Naïve Bayes (92%), which may be influenced by dataset limitations or the model's validation method. Data preprocessing involved several steps, including tokenization, stopword removal, stemming, lemmatization, and handling of missing data to ensure cleaner and more structured text input. These findings highlight the potential of sentiment analysis for transportation policy improvements, providing insights for policymakers and transport service providers. Future research should address data balancing and broader dataset usage to enhance the robustness of findings and support better decision-making in the transportation sector
PRESERVATION OF THROUGH PATTERN RECOGNITION USING A COMBINATION OF GLCM, LBP, AND SVM MULTICLASS
Full text linkThis study proposes an automatic method to recognize traditional Timorese weaving patterns using machine learning techniques. Timorese weaving image data is processed through pre-processing stages and its features are extracted using the Gray Level Cooccurrence Matrix (GLCM) and Local Binary Pattern (LBP) methods, which function to capture the characteristics of texture and design in the weaving patterns. The classification model is built with the Support Vector Machine (SVM) algorithm using the One Versus One (OVO) and One Versus All (OVA) approaches with several kernels, including Linear, Polynomial, and Radial Basis Function (RBF). The best results were obtained with the Linear kernel and the OVO method, resulting in an accuracy of 88.66%, a precision of 88.66%, a recall of 88.80%, and an F1-score of 88.73%. This approach shows great potential in preserving and documenting Timorese weaving patterns automatically and efficiently, with accurate classification results. This study explores a machine learning approach for identifying Timorese weaving patterns. By leveraging GLCM and LBP for texture analysis and SVM with OVO and OVA for classification, the method achieves high accuracy. The findings support digital preservation efforts and cultural heritage conservation
FACIAL RECOGNITION SYSTEM FOR DISTANCE LEARNING STUDENT ATTENDANCE MANAGEMENT USING MACHINE LEARNING
Full text linkThe administration of student attendance constitutes a vital component of academic governance, affecting both educational outcomes and institutional efficacy. Utilizing machine learning to augment precision and efficacy, with adaptability for both physical and remote learning environments. The research methodology encompasses the acquisition of facial data from students under diverse lighting conditions, perspectives, and remote settings, succeeded by preprocessing and training of a facial recognition algorithm employing machine learning techniques. The system addresses key technical challenges such as camera quality variations, lighting inconsistencies, and spoofing risks by integrating robust image preprocessing and security safeguards. Quantitative evaluation shows that under ideal and controlled conditions, the system achieves up to 100% accuracy with an average processing time of 0.8 seconds. With the specifications Intel Core i5, RAM8 GB, minimum windows 10, NVIDIA GeForce GTX 1050, 1080p minimum camera with 30 fps frame rate, standard CMOS sensor, and automatic exposure adjustment capabilities, accuracy will drop if the conditions are not ideal. The system ensures the security and privacy of student facial because it is live with zoom or LMS. Furthermore, the incorporation of this system facilitates the realization of smart campus initiatives by delivering precise, real-time attendance information. This inquiry contributes to educational technology, enhancing operational efficacy and fostering digital transformation within higher education institutions. The designed system also seeks to reduce overall student attendance fraud
PARAMETER TUNING IN BACKPROPAGATION NEURAL NETWORKS: IMPACT OF LEARNING RATE AND MOMENTUM ON PERFORMANCE
Full text linkArtificial Neural Network (ANN) play a pivotal role across diverse domains, including medicine, economics, and technology, due to their ability to model complex relationships and deliver high prediction accuracy. This study systematically examines how learning rate and momentum interact in backpropagation, moving beyond isolated analysis to enhance ANN performance. A qualitative research design employing a systematic literature review was utilized, with data sourced from reputable databases covering the past 11 years. Bibliometric tools such as VOSviewer and R-Studio were applied to identify trends and patterns in the literature. The findings reveal that both learning rate and momentum significantly impact convergence efficiency and model stability. Backpropagation remains fundamental for weight adjustment in minimizing prediction errors. ANN optimization demonstrates substantial practical benefits, including enhanced treatment outcome predictions in medicine, modeling nonlinear patterns in economics, and improved image classification accuracy. However, challenges such as the curse of dimensionality, overfitting, and dependence on large datasets persist. Strategies such as regularization, ensemble methods, and sensitivity analysis present viable solutions. This study underscores the critical need to advance ANN optimization techniques and highlights the potential of interdisciplinary collaboration in addressing existing limitations and broadening ANN application
PREDICTION MODEL OF HUMAN DEVELOPMENT INDEX (HDI) USING K-NEAREST NEIGHBOR (KNN) ENSEMBLE
Full text linkThe Human Development Index (HDI) is an essential indicator in measuring the success of human development. Although some regions in Indonesia have experienced increased HDI, inequality between areas makes it difficult to predict future HDI values. This research aims to build an HDI prediction model using the ensemble K-nearest neighbor (KNN) method. The dataset consists of 574 data points with attributes of life expectancy, expected years of schooling, average years of education, and regional income per capita. The method used is SEMMA with z-score normalization, feature selection based on domain knowledge, and validation with 10-fold cross-validation. The results showed that the KNN Ensemble model with the Boosting (Adaboost) technique had the best performance with an average MAPE of 0.58%, which indicates that the model's predictions deviate by less than 1% from actual HDI values, which is considered highly accurate and reliable for policy planning. This model proved better than linear regression, neural networks, single KNN, and double exponential smoothing algorithms. The improved prediction accuracy of the proposed model provides local governments with a reliable tool for scenario-based development planning and policy simulation, contributing to achieving the Golden Indonesia 2045 strategic vision
COMPARATIVE ANALYSIS OF YOLO DEEP LEARNING MODEL FOR IMAGE-BASED BEEF FRESHNESS DETECTION
Full text linkEnsuring beef freshness is essential to protect consumer health and maintain public trust in the food supply chain. However, conventional freshness assessment relies on subjective human sensory judgment and can be inconsistent. This study presents a comparative evaluation of three YOLO models, YOLOv5sM (with targeted augmentations Flip, Rotation, Mosaic), YOLOv8, and YOLOv11 for automated beef freshness detection in digital images. Unlike prior studies focusing on a single YOLO version, this work systematically compares multiple YOLO generations to assess accuracy and computational efficiency. Evaluation metrics included precision, recall, [email protected], [email protected]:0.95, and training time. A labeled dataset of 4,000 beef images (fresh and non-fresh) was split into training, validation, and test sets, with augmentation applied only to YOLOv5sM. All three models achieved 100% precision and recall on the test set; however, this likely reflects dataset homogeneity and potential overfitting, limiting interpretation of these results. YOLOv11 achieved the highest localization accuracy ([email protected]:0.95 = 97.0%), followed by YOLOv8 (96.9%) and YOLOv5sM (96.2%). YOLOv8 had the shortest training time (54 minutes), whereas YOLOv11 offered the best balance of accuracy, model size (5.4 MB), and computational efficiency. Overall, YOLOv11 emerged as the optimal model, offering superior performance and practical deployment advantages over earlier YOLO versions. As the first systematic comparison of multiple YOLO generations for beef freshness detection, this study provides novel insights into detection accuracy and computational efficiency
SISTEM INFORMASI MANAJEMEN ARSIP PADA DIREKTORAT TEKNOLOGI INFORMASI KEIMIGRASIAN
Full text linkIn this research case study, the storage of letter archives at the Directorate of Immigration Information Technology is still done manually where all letter archive documents are stored by employees on their respective Personal Computers. Records management that is done manually by storing files on Personal Computer rarely has a regular backup mechanism so that if a document is lost it is difficult to recover it. So, an innovation is needed in the form of a website-based archive storage information system. The purpose of this research is to facilitate employees who are appointed as letter archive managers and facilitate the search for archives needed by the leadership. In this study, researchers used the scrum method or model with stages namely product backlog, sprints, scrum meetings and demos. The stages of the Agile method in this study include system analysis, design, development, testing, deployment, system evaluation and maintenance. The programming language used in building archive management information system applications at the directorate of immigration information technology is using the PHP (Hypertext Preprocessor) and JavaScript programming languages. The results showed that the web-based archive management information system has been successfully designed. This system overcomes difficulties in searching for archives, reduces the risk of data loss, and optimizes the management of incoming and outgoing mail archives. With this system, officers in each section can manage the storage of letter archives and enable data management and document searches that were previously time-consuming now become faster and more efficient