ejournal.nusamandiri.ac.id (STMIK Nusa Mandiri)
Not a member yet
1504 research outputs found
Sort by
PERFORMANCE EVALUATION OF NEWTON–KONTOROVICH AND ADAPTIVE NEWTON LINE SEARCH ON MULTIVARIATE NONLINEAR SYSTEMS
Full text linkSolving multivariate nonlinear systems is essential in engineering, physics, and applied sciences. This study compares the performance of two numerical methods—Newton–Kontorovich and Interactive Newton–Raphson with Line Search—on trigonometric and exponential nonlinear systems. The methods are evaluated based on convergence rate, accuracy, and iteration efficiency through numerical simulations using MATLAB. The Newton–Kontorovich method, typically used for integral or differential equations, is compared with the adaptive line search strategy that enhances global convergence. Results show that the Interactive Newton–Raphson method achieves a smaller final error (5.95×10⁻²) with stable convergence, while Newton–Kontorovich converges in fewer iterations but with larger error (3.126). These findings highlight the superiority of adaptive strategies for complex nonlinear systems. Practical implications include improved numerical reliability for applications in structural engineering, optimization, and scientific modeling
QUALITY ANALYSIS OF THE ELECTRONIC GOVERNMENT PROCUREMENT ORDER SYSTEM USING WEBQUAL AND EUCS METHODS
Full text linkWithin district of Hulu Sungai Selatan there is a website for the community to offer goods or services, namely SIOPEN, SIOPEN functions as a forum for the HSS community and umkm to offer their goods or services to the local government. The goal of this research is to evaluate SIOPEN's quality using the Webqual approach and users' degree of satisfaction using the EUCS approach. The Webqual approach, which has three dimensions information quality, usability, and interaction service was employed in this research together with the EUCS approach. The EUCS approach has some variables in partikular ease of use, accuracy, format, content, and timeliness. The approach of gathering information data through respondents via sending out online questionnaires to SIOPEN users and determining the sample using the formula from Slovin obtained 88 users then obtained by Webqual Index and Average Satisfaction. The results of research measuring the three dimensions of Webqual show that are in very good interpretation, The information quality dimension obtained 0.8392, the usability dimension obtained 0.8292, then the service interaction dimension obtained 0.8284. The findings from assessing the level of the satisfied user three variables of EUCS in particular content at a score of 4.27, ease of use at a score of 4.21, and timeliness at a score of 4.28 are at very satisfied level, then the other two variables are accuracy at a score of 4.15 and format at a score of 4.11 at a satisfied level
ONLINE DELIVERY, DINE-IN, AND RESERVATION SYSTEM USING THROW-AWAY PROTOTYPING AT JONG JAVA RESTAURANT
Full text linkThe culinary industry continues to grow amidst the increasing purchasing power and consumption of society. Many restaurants still rely on direct sales (dine-in) without utilizing technology to enhance services and expand their market. This study aims to design and build a web-based information system for Restoran Jong Java, which includes dine-in, online delivery, and table reservation features. The system is expected to simplify transaction processes, improve operational efficiency, and provide added value for the restaurant in the industry competition. The system is developed using the throw-away prototyping method. This method allows for iterative system development with direct input from users. Testing is conducted using black-box testing to ensure the system functions as per the specifications, and user acceptance testing is performed through questionnaires with five main types of users: customers, admins, cashiers, waiters, and waitresses. The designed system is capable of supporting the integrated management of dine-in services, reservations, and online deliveries. The tests show that the system meets user needs, with a high level of satisfaction from the respondents. The user acceptance testing in this study shows positive results across different user groups. For the customer group, the average score obtained was 77.25%, the waiter group gave an average score of 83.6%, and for the managers, the system was also well received by them. This system has successfully improved the restaurant's operational efficiency and provided convenience for customers in making orders. It also serves as a technological solution that can help Jong Java expand its market reach and increase competitiveness
STUDENT ATTENDANCE BASED ON FACE RECOGNITION USING THE CONVOLUTIONAL NEURAL NETWORK METHOD
Full text linkMataram University of Technology (UTM) still relies on a manual attendance process, such as signing paper-based attendance lists, which are prone to fraud and difficult to manage on a large scale. This study develops a face recognition-based attendance system using Convolutional Neural Network (CNN), which can automatically recognize visual patterns and unique facial features. CNN has advantages in extracting significant facial features, allowing it to recognize faces under various lighting conditions and viewing angles. The dataset used consists of 5,820 facial images from 97 students, with 60 augmented images per student. The results indicate that this system can be implemented in a lecture environment, achieving a validation accuracy of 98.5% at the 150th epoch. However, the model has some limitations, such as a relatively small dataset size and challenges in recognizing faces under extreme lighting conditions or unusual angles, which can affect accuracy in real-world applications. Additionally, although this system has the potential for real-time implementation, further optimization is required to ensure fast and accurate responses on a large scale. To overcome these limitations, future research can explore the use of direct camera input to enhance efficiency and user experience. Furthermore, improving dataset quality by incorporating variations in lighting and image angles, as well as exploring alternative deep learning architectures such as Vision Transformers (ViT) or Swin Transformer, can enhance model performance and generalization. By implementing these improvements, the facial recognition-based attendance system can be more optimal in enhancing accuracy and ease of use in academic environments
PREDICTION OF HAJJ PILGRIMS' HEALTH RISK USING K-NN, DECISION TREE, CROSS VALIDATION, AND SMOTE
Full text linkThe background of this study is predicting the health risk levels of hajj pilgrims, which is a significant challenge in improving healthcare services during the pilgrimage. This research contributes by systematically evaluating several machine learning techniques and applying SMOTE to balance the dataset, as opposed to previous studies that relied on single-model classification approaches. The data analyzed includes 5,000 health records of pilgrims, covering various attributes such as age, gender, medical history, and disease diagnosis, sourced from the Siskohat database of the Directorate General of Hajj and Umrah Management. The results show that Cross-Validation (Logistic Regression) achieved the highest accuracy (87.9%) after applying SMOTE, outperforming Decision Tree (86.4%) and K-NN (83.1%). These findings highlight that SMOTE significantly enhances recall, ensuring better identification of high-risk patients. The implications of these results contribute to hajj health management by providing a robust predictive framework that improves early risk detection and medical resource allocation, while also demonstrating a novel approach to handling imbalanced healthcare datasets
EARLY DETECTION OF ROT IN THAI PAPAYA (CARICA PAPAYA) USING THE K-NN METHOD
Full text linkDetermining the category of a plant or fruit involves several criteria. One of the easiest methods to use is morphological criteria, which entails studying the external structure that can be directly observed. However, this approach cannot be regarded as a fixed standard since people's interpretations may vary. To address this, a system was developed to assess the ripeness of Thai papaya fruit, utilizing image processing and the K-Nearest Neighbor (KNN) method. This study analyzes a data set to detect rotten papaya fruit, which is expected to help consumers recognize papaya fruit that is purchased in a perfectly ripe condition, not ripe with certain parts that are rotting. The indicator used to determine the category is the color of the skin of Thai Papaya fruit with an ROI of 600 pixels x 300 pixels by finding the mean RGB value and then calculating it using the Euclidean distance formula. From the results of these calculations, it is expected to get a classification using K-Nearest Neighbor (KNN) to get an image pattern of the level of rottenness on the surface of the papaya. Therefore, by improving the RGB image eliminating noise in the papaya image, and using the K-NN classification of the image pattern obtained from the research results from the sampling data, an accuracy level of 80% was obtained with a range of mean R values: 130,671-169,630, mean G: 106,891-131,895, and mean B: 61,119-100,776 which came from 120 data
OPTIMIZING THE KNN ALGORITHM FOR CLASSIFYING CHRONIC KIDNEY DISEASE USING GRIDSEARCHCV
Full text linkChronic Kidney Disease (CKD) is a progressive condition that impairs kidney function and cannot be cured. Early detection is crucial for effective management and therapy. However, diagnosing CKD is challenging as patients often have comorbidities such as diabetes, hypertension, or heart disease, which complicate diagnosis and treatment. Accurate classification methods are essential for early detection. K-Nearest Neighbor (KNN) is a classification algorithm that groups data based on feature similarity. K-NN is an algorithm that is resistant to outliers, easy to implement, and highly adaptable. It only requires distance calculations between data points and does not involve complex parameters. However, its performance depends on hyperparameters such as the number of neighbors (k), weighting, and distance metric. Incorrect hyperparameter selection can lead to overfitting, underfitting, or reduced accuracy. To address these issues, GridSearchCV is used to optimize KNN by systematically selecting the best hyperparameters, ensuring improved accuracy and reduced overfitting. This optimization enhances the model’s reliability in early CKD detection compared to other methods. This study aims to determine the optimal KNN parameters for CKD classification using GridSearchCV. The results show 8.05% accuracy improvement and reduction in overfitting, with the prediction gap between training and testing decreasing from 6% to only 1.15%. These enhancements contribute to more reliable CKD diagnosis, enabling accurate early detection and better clinical decision-making
IMPLEMENTATION OF CNN FOR CLASSIFYING PATCHOULI LEAF IMAGES BASED ON ACCURACY AND EVALUATION
Full text linkPatchouli (Nilam leaves) holds significant potential as a high-value natural material, especially in the perfume and essential oil industries. However, the classification and quality analysis of patchouli leaves remain a challenge that requires an automated solution based on technology. This study aims to develop a Convolutional Neural Network (CNN) model capable of automatically classifying the condition of patchouli leaves. The image data of patchouli leaves were processed through several preprocessing stages and divided into training and testing data. The designed CNN model utilizes several convolutional layers, pooling, dropout, and dense layers for the training process. The evaluation results using the confusion matrix showed that the model had a very low error rate, with only 1 misprediction in the training data. For the testing data, the model achieved an accuracy of 85% with a loss value of 0.6191496. The model also demonstrated an accuracy of 98.75% with a loss of 0.443462 on the training data. However, improvements in model generalization are still needed to achieve more consistent performance on new dat
PWA AND NON-PWA PERFORMANCE ANALYSIS: CHROME EXTENSION TESTING ON E-COMMERCE PLATFRORM
Full text linkThis study compares Progressive Web Apps (PWA) and traditional web applications performance using a custom Chrome extension and Google Lighthouse, focusing on Tokopedia's e-commerce platform. The research employs a quantitative approach with controlled testing environments across three viewports for the custom extension (desktop, tablet, mobile) and two viewports for Google Lighthouse (desktop, mobile). The custom extension measures eleven metrics, including Core Web Vitals, PWA features, and resource usage, while Google Lighthouse provides five core metrics. Results show PWA implementation improves performance with 9.9% better First Contentful Paint on desktop and significant memory efficiency (29-33MB vs 59-62MB). The comparison between testing tools reveals methodology differences, with custom extension showing optimistic results in real-world conditions and Lighthouse providing more conservative measurements under throttled conditions. This research contributes to PWA performance measurement methodology by combining real-world and standardized testing approaches
COMPARATIVE OF LSTM AND GRU FOR TRAFFIC PREDICTION AT ADIPURA INTERSECTION, BANDAR LAMPUNG
Full text linkThe Tugu Adipura intersection in Bandar Lampung is a vital traffic hub connecting four major roads. Rapid population growth and increasing vehicle numbers challenge traffic flow and urban quality of life. Despite its importance, there is limited research using predictive models to analyze traffic patterns at complex intersections in mid-sized Indonesian cities. This study addresses that gap by examining traffic growth on four connected roads using deep learning models. Traffic data were collected hourly from June 1, 2021, to July 31, 2023. A comparative analysis of Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) models was conducted, with SGD and Adam as optimizers. Results show the GRU model with Adam achieved the lowest RMSE (0.23) on road section 1, indicating its superior ability to model short-term fluctuations and non-linear growth in traffic volume. The study offers practical implications for traffic management by highlighting GRU’s capacity to capture seasonal trends and rapid growth, supporting proactive infrastructure planning and congestion mitigation strategies. These findings demonstrate the value of data-driven approaches in enhancing transportation systems in growing urban areas