ejournal.nusamandiri.ac.id (STMIK Nusa Mandiri)
Not a member yet
1504 research outputs found
Sort by
OPTIMIZATION OF SVM ALGORITHM FOR OBESITY CLASSIFICATION WITH SMOTE TECHNIQUE AND HYPERPARAMETER TUNING
Get PDFExcessive fat accumulation that impairs personal health and raises the risk of chronic diseases is the hallmark of obesity, a global health issue. Decision Tree (DT) has been widely used for obesity classification, but it tends to suffer from overfitting and poor performance on imbalanced datasets. To overcome these limitations, this study proposes an optimization of the Support Vector Machine (SVM) algorithm using Synthetic Minority Over-sampling Technique (SMOTE) and Hyperparameter Tuning. SMOTE was applied to balance the class distribution, whereas Grid Search was utilized to determine the optimal combination of hyperparameters (C, gamma, and kernel). The dataset employed in this research comprises multiple features related to individual health and lifestyle, with obesity level as the target class. The experimental results demonstrate that the optimized SVM model demonstrated strong classification performance, attaining 97% in accuracy, precision, recall, and F1-score. This high performance is significant because it enables more accurate early detection of obesity risk, which can support timely medical intervention and personalized treatment planning, ultimately contributing to better public health outcomesThese findings suggest that incorporating SMOTE and Hyperparameter Tuning substantially improves SVM performance, establishing it as a robust approach for obesity classification and early risk detection
DEVELOPMENT OF A SMART IOT-BASED MONITORING SYSTEM FOR FERTIGATION AND SEED WEIGHT DETECTION IN SACHA INCHI
Get PDFThis research focuses on designing a fertilization monitoring system based on the Internet of Things (IoT) and detecting the weight of Sacha Inchi plant seeds. The two tools are integrated with IoT platforms, enabling remote monitoring and control via the Simosachi app. Test results indicate that the system provides accurate data on soil and plant conditions, allowing farmers to make informed decisions on fertilization and irrigation. The seed weight detection tool also functions well, with a minor error margin still within acceptable limits. With improved monitoring and control of the fertilization process, as well as accurate monitoring of crop yields, the system is expected to help farmers achieve more optimal harvests. The seed weight detection tool achieved an accuracy of 97.94%, surpassing similar prior systems in terms of real-time data integration and multi-parameter monitoring. Future research may focus on enhancing the accuracy of the seed weight detection tool and developing advanced fertigation control algorithm
APPLICATION OF RANDOM FOREST ALGORITHM FOR ARRHYTHMIA DETECTION BASED ON ELECTROCARDIOGRAM DATA
Get PDFArrhythmia is a common cardiac disorder that requires early detection to prevent serious complications. This study applied the Random Forest algorithm to enhance electrocardiogram (ECG) analysis and enable accurate arrhythmia classification. Unlike prior studies that focused primarily on resting ECG signals, this research incorporated dynamic data collected from 26 participants performing three physical activities for three minutes each, capturing physiological variations across multiple activity states. The Random Forest model was constructed and evaluated using ECG-derived temporal and morphological features to detect potential arrhythmias. Experimental results showed that the model achieved an accuracy of 97.4%, with precision, recall, and F1-score each reaching 98%, and an AUC of 0.97. However, several limitations remain, including the relatively small and homogeneous sample, as well as the short recording duration. Nonetheless, the proposed approach demonstrates strong potential to support early cardiac screening and real-time monitoring, particularly in portable and resource-limited healthcare application
PENDAMPINGAN MASYARAKAT KELURAHAN PARIT LALANG DALAM DIGITALISASI DAN SPASIALISASI WILAYAH
Get PDFSpatial planning requires diverse, comprehensive, accurate, and detailed data to forecast spatial development needs. Parit Lalang Subdistrict is located in Rangkui District, Pangkal Pinang City. The potential of Parit Lalang Subdistrict has yet to be identified well enough and accurately. Community assistance aims to digitize and spatialize the potential of Parit Lalang Village. This community program is carried out in three stages: preparation, implementation, and evaluation. The team carried out coordination activities and prepared training modules in the preparation stage. Participatory mapping activities (spatialization) and spatial data digitization training were carried out at the implementation stage. At the evaluation stage, evaluation meetings are carried out, and the results of the activities are submitted to partners. This community assistance activity results in data on potential and neighborhood problems that can be mapped and published digitally as a database for preparing planning programs. The database is arranged in map form, including regional boundary maps, land use maps, and supporting infrastructure maps. From this mentoring activity, a digitalization and spatialization module was also produced, a guidebook that can be used by village governments if they want to map regional potential. The enthusiasm of the people and government of Parit Lalang Subdistrict is quite high in identifying the area's potential so that similar activities can be replicated in other villages. The availability of digital spatial data formats can increase the reach of public information on the potential and conditions of residential areas, thereby providing efficiency in meeting community needs from a development aspect
EDUKASI PERLINDUNGAN DATA PRIBADI DALAM LAYANAN PUBLIK PADA MASYRAKAT DAERAH PEMILIHAN PAPUA
Get PDFThe protection of personal data has become a critical concern in Indonesia, particularly with the increasing cases of data breaches involving healthcare platforms and government institutions. In public services, safeguarding personal data is vital to maintaining citizens’ trust in government systems. This community service program aims to enhance public and stakeholder awareness of regulations, digital technologies, and best practices in personal data protection. Activities were carried out through an educational approach, including seminars and hands-on training. The materials covered the urgency of personal data protection, relevant Indonesian regulations such as Law No. 27 of 2022, and practical digital security techniques such as two-factor authentication, full encryption, and safe internet practices. The results indicated a significant improvement in participants’ understanding of personal data protection and their ability to apply digital security measures. Participants also received technical guidance on the use of emerging technologies, including biometric authentication and secure password management. In conclusion, effective personal data protection in public services requires not only strong legal frameworks but also continuous public engagement through education and technology adoption. This program contributes to building a safer and more trustworthy digital ecosystem in Indonesia and serves as a replicable model for enhancing digital literacy in public institutions
DESIGNING A MICROSERVICES BASED ENTERPRISE ARCHITECTURE USING TOGAF 10: A CASE STUDY APPROACH
Get PDFPT Umbara Karya Sejati, operating in the home living and industrial product sectors, faced significant digital transformation challenges, marked by inefficiencies due to system integration issues and vendor dependencies that incurred cost penalties of 10-15% each month. Addressing these challenges, this research developed an Enterprise Architecture using TOGAF 10's Microservices Architecture (MSA) from the Preliminary Phase to Phase F: Migration Planning. This approach aimed to enhance system integration and operational efficiency, thereby improving modularity and scalability while reducing reliance on external vendors. A pivotal component of this architecture, an API gateway, provided robust monitoring capabilities for integration processes, enabling quick identification and prioritization of critical issues, which will assist the internal IT team's workflow. Furthermore, the research planned the establishment of a DevOps team, incorporating Agile methodologies, and scheduled IT governance and data security training to prepare for future policy development and strengthen internal controls. This strategic design equips the organization to navigate the volatility, uncertainty, complexity, and ambiguity (VUCA) of market demands with agile and effective responses, projecting a 10% increase in both operational and cost efficiency for PT Umbara Karya Sejati
IMPLEMENTASI HYBRID INTELLIGENCE SYSTEM UNTUK KLASIFIKASI BIJI-BIJIAN DENGAN ALGORITMA PCA DAN KNN
Get PDFFood security has become a pressing global issue with the increasing population and food consumption needs. Red kidney beans, peanuts, and sunflower seeds play a crucial role in meeting the nutritional needs of society and serving as raw materials for various industries. This study aims to develop a seed classification system based on the Principal Component Analysis (PCA) and K-Nearest Neighbor (KNN) algorithms. The system is designed to recognize three types of seeds—red kidney beans, peanuts, and sunflower seeds—to improve the efficiency and accuracy of the classification process compared to manual methods. The dataset consists of 58 seed image samples, divided into training data (48 samples) and test data (10 samples). The research stages include image preprocessing (cropping, background removal, and thresholding segmentation), feature extraction using PCA to reduce data dimensionality, and classification with KNN based on Euclidean distance. A value of K=3 is used in the KNN algorithm to determine the proximity between data points. The test results show a classification accuracy of 90%, with 9 out of 10 test data correctly classified. PCA successfully simplified high-dimensional data into two main components without significant information loss, while KNN demonstrated strong capability in distinguishing the three types of seeds. This research contributes to the development of an AI-based automatic classification system for the food industry, with broader potential applications in high-dimensional data processing across various fields
PROTOTIPE KERAN AIR TANPA SENTUH DAN PENGUKUR SUHU TUBUH OTOMATIS BERBASIS MIKROKONTROLLER ARDUINO UNO
Get PDFThe global pandemic requires the development of technological solutions to minimize physical contact, especially in public facilities such as water taps, which can function as a medium for transmitting infectious diseases. This study aims to design and develop a prototype of an automatic water tap integrated with an Arduino Uno microcontroller-based body temperature meter. This system was created to support health protocol efforts when carrying out activities, increasing efficiency and reducing the risk of disease transmission. The research method includes problem identification, literature study, hardware and software component design, prototyping, and functionality testing. The test results obtained show that all components work according to their functions with a high level of accuracy, such as the HC-SR04 ultrasonic sensor, which is able to detect objects at a distance between the object and the sensor <12cm, then the Relay will be active and the Mini Water Pump will pump water automatically, and the Valve on the Solenoid Valve will open, and water will flow automatically through the water tap. The test results on the MLX-90614 temperature sensor also obtained an average difference of only 0.28 ° C compared to the thermometer gun as a comparison
ECG-BASED ARRHYTHMIA DETECTION USING THE NARROW NEURAL NETWORK CLASSIFIER
Get PDFElectrocardiograms (ECG) are important for detecting arrhythmias. Conventional models such as CNN and LSTM are accurate but require large amounts of computation, making them difficult to use on wearable devices and for real-time monitoring. This study evaluates the Narrow Neural Network Classifier (NNNC) as a lightweight and efficient alternative. The dataset consists of 21 subjects with 881 ECG samples, categorized based on walking, sitting, and running activities, and processed through bandpass filtering, normalization, and P-QRS- T wave segmentation. The data is divided into training (70%), validation (15%), and test (15%) sets. The NNNC has 11 convolutional layers, a ReLU activation function, a Softmax output, and 120,000 parameters. The model was trained using the Adam optimizer, a batch size of 32, and a learning rate of 0.001 for 100 epochs and compared with SVM, CNN, and LSTM using accuracy, precision, recall, F1-score, and ROC-AUC. The results show that NNNC achieves an accuracy of 98.9%, a precision of 99.2%, a recall of 99.2%, and an F1-score of 99.2%, higher than SVM and comparable to CNN/LSTM, with lower computational consumption. The model is capable of reliably detecting early arrhythmias. These findings support the potential of NNNC for ECG-based automatic diagnostic systems, including real-time implementation on wearable devices, although further research is needed for large-scale validatio
OPTIMIZATION OF EFFICIENTNET-B0 ARCHITECTURE TO IMPROVE THE ACCURACY OF GLAUCOMA DISEASE CLASSIFICATION
Get PDFGlaucoma is a chronic eye disease that can potentially cause permanent blindness if not detected early. This study aims to improve the generalization capability and reliability of glaucoma classification by optimizing the EfficientNetB0 architecture based on a Convolutional Neural Network (CNN). Optimization was carried out by applying double dropout (0.4 and 0.3) and adding a Dense layer with 128 ReLU-activated neurons to reduce overfitting and strengthen non-linear feature representation. The dataset used consists of 1,450 fundus images (899 glaucoma and 551 normal) obtained from IEEE DataPort. Model performance evaluation was performed using accuracy, precision, recall (sensitivity), specificity, F1 score, and Area Under the Curve (AUC) metrics, complemented by confusion matrix analysis to assess overall classification performance. The results showed that the optimized EfficientNetB0 model consistently outperformed the baseline comparison model with the highest accuracy, precision, recall (sensitivity), specificity, F1 score, and AUC values of 95%. Based on the system performance results obtained, the Proposed model can be used as an aid for medical personnel in classifying glaucoma conditions so that they can provide appropriate medical treatment and reduce the risk of permanent blindness due to glaucoma