Technema: Journal of Intelligent Engineering and Computing
Technema: Journal of Intelligent Engineering and Computing is a peer-reviewed academic journal dedicated to publishing high-quality scholarly work in the fields of engineering, computer science, and emerging technologies. The journal provides an international platform for researchers, engineers, scientists, and technology practitioners to disseminate original research articles, theoretical contributions, experimental studies, system designs, and applied innovations that address contemporary challenges in intelligent systems and digital transformation. Technema welcomes manuscripts in areas such as artificial intelligence, machine learning, data science, computer engineering, software engineering, information systems, cybersecurity, robotics, automation, electrical engineering, the Internet of Things, cloud computing, embedded systems, smart infrastructure, and interdisciplinary technological innovation. All submissions undergo a rigorous double-blind peer-review process to ensure originality, technical rigor, scientific validity, and meaningful contributions to engineering and computing knowledge. Published quarterly in March, June, September, and December, Technema aims to foster global technological dialogue, promote cutting-edge research, and strengthen the role of intelligent engineering and computing solutions in advancing industry, infrastructure, and digital society at local, national, and international levels.
Articles
10 Documents
<![CDATA[Short-Term Electricity Load Forecasting Using Hybrid CNN–LSTM Models on the UCI Electricity Load Dataset ]]>
<![CDATA[Ricky Imanuel Ndaumanu]]>;
<![CDATA[Alfred Yulius Arthadi Putra]]>
<![CDATA[ Technema: Journal of Intelligent Engineering and Computing Vol. 1 No. 1 (2026): March: Technema: Journal of Intelligent Engineering and Computing ]]>
Publisher : <![CDATA[CV SCRIPTA INTELEKTUAL MANDIRI ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
<![CDATA[Accurate short-term electricity load forecasting (STLF) is critical for efficient energy management, demand response, and grid stability in modern residential environments. This study presents an empirical investigation of a hybrid convolutional neural network–long short-term memory (CNN–LSTM) model applied to the UCI Electricity Load dataset, integrating convolutional layers to extract localized temporal features and stacked LSTM layers to model long-term dependencies across households. The model is trained using Python and TensorFlow on a GPU-enabled workstation, with preprocessing including normalization, sliding-window sequence generation, and train-validation-test splitting. Performance is evaluated through mean absolute error (MAE), root mean square error (RMSE), mean absolute percentage error (MAPE), and k-fold cross-validation. Comparative benchmarking against LSTM, CNN-GRU, and multi-scale CNN-LSTM architectures demonstrates superior accuracy, stability, and generalizability of the proposed hybrid model. Sensitivity and interpretability analyses further reveal critical temporal patterns, feature contributions, and operational insights, facilitating actionable energy management decisions. These results substantiate the hybrid CNN–LSTM approach as a robust, interpretable, and operationally relevant solution for STLF applications.]]>
<![CDATA[Network Intrusion Detection Using Ensemble Learning Techniques on the CIC-IDS2017 Public Dataset ]]>
<![CDATA[Sugeng Hendra Wijaya]]>;
<![CDATA[Andhika Adnan]]>
<![CDATA[ Technema: Journal of Intelligent Engineering and Computing Vol. 1 No. 1 (2026): March: Technema: Journal of Intelligent Engineering and Computing ]]>
Publisher : <![CDATA[CV SCRIPTA INTELEKTUAL MANDIRI ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
<![CDATA[Network intrusion detection remains a critical challenge in cybersecurity due to evolving attack patterns, class imbalance, and high-dimensional network traffic data. This study investigates the effectiveness of ensemble learning techniques on the CIC-IDS2017 public dataset, integrating decision trees, random forests, and gradient boosting models through stacking, voting, and hybrid Boost-Bag strategies. Data preprocessing involved normalization, handling missing values, and feature selection based on correlation and mutual information to reduce dimensionality while preserving predictive relevance. Empirical evaluation employed stratified 10-fold cross-validation and performance metrics including accuracy, recall, F1-score, and AUC-ROC, with additional analyses of confusion matrices and temporal stability to assess operational reliability. Results indicate that hybrid ensembles achieve superior detection performance, particularly for low-frequency attacks, while maintaining moderate computational overhead compared to individual classifiers. Comparative insights reveal trade-offs between accuracy, minority-class sensitivity, and inference latency, guiding practical deployment considerations. The findings substantiate the theoretical benefits of ensemble diversity and optimized feature selection, offering a robust framework for scalable, interpretable, and resilient network intrusion detection systems.]]>
<![CDATA[Air Quality Index Prediction Using Machine Learning Algorithms on the Beijing PM2.5 Dataset ]]>
<![CDATA[Tuti Susilawati]]>;
<![CDATA[Bustomi Bustomi]]>
<![CDATA[ Technema: Journal of Intelligent Engineering and Computing Vol. 1 No. 1 (2026): March: Technema: Journal of Intelligent Engineering and Computing ]]>
Publisher : <![CDATA[CV SCRIPTA INTELEKTUAL MANDIRI ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
<![CDATA[Accurate prediction of Air Quality Index (AQI) is critical for mitigating public health risks associated with urban air pollution. This study presents an empirical analysis of PM2.5 concentration forecasting in Beijing using advanced machine learning algorithms, integrating high-resolution atmospheric data and meteorological variables. A multi-stage pipeline was implemented, including data preprocessing, feature selection, and model training with Random Forest, Gradient Boosting, Support Vector Regression, and Long Short-Term Memory (LSTM) networks. Predictive performance was evaluated using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and R², while spatial and temporal fidelity was assessed across districts, diurnal cycles, and seasonal periods. LSTM models consistently achieved superior accuracy, capturing short-term pollution spikes, seasonal variability, and spatial heterogeneity, whereas ensemble methods provided stable baseline predictions with moderate sensitivity to extreme events. Sensitivity analysis identified wind speed, humidity, and neighboring PM2.5 measurements as key predictors. The results demonstrate that integrating recurrent neural networks with ensemble approaches enables reliable, operationally relevant AQI forecasts, offering both theoretical validation of sequential modeling for urban air quality and practical guidance for environmental monitoring, public health interventions, and city-level policy implementation. ]]>
<![CDATA[Credit Card Fraud Detection Using Random Forest and XGBoost on a Public Kaggle Dataset ]]>
<![CDATA[Erlita Sulistiati]]>;
<![CDATA[Tuti Susilawati]]>
<![CDATA[ Technema: Journal of Intelligent Engineering and Computing Vol. 1 No. 1 (2026): March: Technema: Journal of Intelligent Engineering and Computing ]]>
Publisher : <![CDATA[CV SCRIPTA INTELEKTUAL MANDIRI ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
<![CDATA[Credit card fraud detection remains a critical challenge in digital financial ecosystems characterized by extreme class imbalance and evolving attack patterns. This study adopts an empirical experimental design to evaluate and compare Random Forest and XGBoost models using the publicly available Kaggle Credit Card Fraud Detection dataset. A controlled pipeline incorporating stratified data splitting, SMOTE-based imbalance mitigation, and GridSearchCV hyperparameter optimization was implemented to ensure methodological consistency and reproducibility. Performance was assessed through precision, recall, F1-score, ROC-AUC, confusion matrix analysis, and computational efficiency metrics. Results indicate that XGBoost outperformed Random Forest in recall, F1-score, and ROC-AUC, demonstrating enhanced minority-class discrimination and reduced false negatives under optimized conditions. Random Forest exhibited competitive precision and interpretability transparency, though with slightly lower sensitivity. Scalability evaluation confirmed that both models maintained low inference latency suitable for near-real-time deployment. The findings highlight the critical role of imbalance handling and parameter optimization in ensemble-based fraud detection and support boosting-oriented approaches as strategically advantageous for operational financial security systems. ]]>
<![CDATA[Human Activity Recognition Using Deep Learning Models on the UCI HAR Dataset ]]>
<![CDATA[Manorang Gultom]]>;
<![CDATA[Nurhikmah Nurhikmah]]>
<![CDATA[ Technema: Journal of Intelligent Engineering and Computing Vol. 1 No. 1 (2026): March: Technema: Journal of Intelligent Engineering and Computing ]]>
Publisher : <![CDATA[CV SCRIPTA INTELEKTUAL MANDIRI ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
<![CDATA[Human Activity Recognition (HAR) has emerged as a critical component of wearable sensing and intelligent healthcare systems, necessitating robust and computationally efficient deep learning architectures. This study presents an empirical experimental evaluation of three deep learning models 1D Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and a hybrid CNN–LSTM using the UCI HAR dataset under a subject-independent protocol. Raw inertial signals were segmented, normalized, and processed through standardized preprocessing pipelines to ensure reproducibility. Performance was assessed using accuracy, precision, recall, F1-score, cross-validation stability, and computational efficiency metrics. Results indicate that the CNN–LSTM architecture achieves the highest test accuracy (94.87%) and demonstrates improved robustness with lower variance and reduced sensitivity to signal perturbations compared to standalone models. Computational analysis confirms that the hybrid configuration maintains feasible inference latency for real-time applications despite moderate increases in parameter size. The findings validate the effectiveness of integrated spatiotemporal feature learning and provide a reproducible benchmark for future research on deep learning–based HAR systems in wearable and IoT contexts.]]>
<![CDATA[Explainable Artificial Intelligence for Engineering Decision Support Systems ]]>
<![CDATA[Rezi Munizar]]>;
<![CDATA[Ahmad Budi Trisnawan]]>;
<![CDATA[Maiza Fikri]]>;
<![CDATA[Tarma Tarma]]>;
<![CDATA[Eko Sutrisno]]>
<![CDATA[ Technema: Journal of Intelligent Engineering and Computing Vol. 1 No. 2 (2026): : June: Technema: Journal of Intelligent Engineering and Computing ]]>
Publisher : <![CDATA[CV SCRIPTA INTELEKTUAL MANDIRI ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
<![CDATA[Explainable Artificial Intelligence (XAI) has become increasingly important in Engineering Decision Support Systems because growing algorithmic complexity often reduces transparency, accountability, and practitioner confidence despite substantial improvements in predictive capability. This study develops a conceptual framework that integrates engineering data processing, artificial intelligence inference, explainability mechanisms, and human-centered decision support within a unified multilayer architecture. A non-empirical system design methodology based on design-oriented systems research, architecture-based evaluation, and scenario-driven analytical simulation is employed to examine architectural consistency, interoperability, traceability, and explainability across representative engineering contexts. The analytical results indicate that explanation fidelity, interpretability, transparency, traceability, and modular scalability function as complementary engineering quality attributes that collectively strengthen trustworthy decision support while preserving logical consistency between engineering evidence and computational reasoning. The proposed architecture also demonstrates technology independence and adaptability across heterogeneous engineering domains through explicit modular interactions and standardized information flows. This study contributes an architecture-centered perspective that advances theoretical understanding of explainable engineering intelligence while providing a reproducible conceptual foundation for future empirical implementation, quantitative validation, and standardized evaluation of trustworthy Engineering Decision Support Systems. ]]>
<![CDATA[Digital Twin Technology for Sustainable Industrial Operations ]]>
<![CDATA[Erlita Sulistiati]]>;
<![CDATA[Bustomi Bustomi]]>;
<![CDATA[Guslila Sari Nasution]]>;
<![CDATA[Atina Salamah]]>;
<![CDATA[Rian Ardianto]]>
<![CDATA[ Technema: Journal of Intelligent Engineering and Computing Vol. 1 No. 2 (2026): : June: Technema: Journal of Intelligent Engineering and Computing ]]>
Publisher : <![CDATA[CV SCRIPTA INTELEKTUAL MANDIRI ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
<![CDATA[Digital Twin technology has emerged as a strategic enabler for sustainable industrial transformation by integrating physical operations, virtual representations, predictive analytics, and sustainability-oriented decision support into a unified cyber–physical environment. This study aims to develop and analytically evaluate a comprehensive Digital Twin framework capable of supporting sustainable industrial operations through the integration of operational efficiency, energy performance, resource optimization, and system resilience dimensions. A non-empirical system design approach was employed to construct a multilayer architecture consisting of physical operation, data acquisition, communication and synchronization, digital twin modeling, analytics and optimization, and sustainability decision-support layers. Technical evaluation was conducted through model-based simulation and analytical assessment using standardized sustainability and operational indicators. The findings demonstrate that the proposed framework strengthens operational visibility, predictive maintenance capability, energy efficiency, resource utilization, responsiveness, and resilience through continuous interaction between physical and virtual environments. The analysis further indicates that Digital Twin integration facilitates circularity, sustainability governance, and Industry 5.0 readiness by enabling adaptive and data-driven industrial decision making. The study contributes a holistic conceptual framework that advances the understanding of Digital Twin technology as a sustainability-enabling infrastructure for future industrial systems. ]]>
<![CDATA[Edge Intelligence in Smart Manufacturing Ecosystems ]]>
<![CDATA[Irwandi Rizki Putra]]>;
<![CDATA[Anjela Karunia Amalia]]>;
<![CDATA[Krisna Widi Nugraha]]>;
<![CDATA[Tuti Susilawati]]>;
<![CDATA[Hadi Jayusman]]>
<![CDATA[ Technema: Journal of Intelligent Engineering and Computing Vol. 1 No. 2 (2026): : June: Technema: Journal of Intelligent Engineering and Computing ]]>
Publisher : <![CDATA[CV SCRIPTA INTELEKTUAL MANDIRI ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
<![CDATA[The rapid evolution of smart manufacturing ecosystems has intensified the need for intelligent architectures capable of supporting real-time decision-making, operational resilience, and sustainable industrial performance. This study investigates the effectiveness of Edge Intelligence within smart manufacturing environments through an empirical system-design and experimental validation approach. A three-layer architecture consisting of the Industrial Internet of Things layer, the edge intelligence layer, and the cloud orchestration layer was developed and evaluated under predictive maintenance, production scheduling, and anomaly detection scenarios. Performance assessment employed metrics including inference latency, response time, bandwidth consumption, prediction accuracy, throughput, resource utilization, reliability, resilience, and energy efficiency. The experimental results demonstrate that edge-enabled intelligence significantly improves manufacturing performance by reducing latency and communication overhead while increasing operational responsiveness, decision consistency, throughput, and system reliability. The architecture also enhances adaptive decision-making capabilities, strengthens human-machine collaboration, improves cybersecurity resilience, and contributes to environmental sustainability through more efficient resource utilization and reduced carbon emissions. The findings establish Edge Intelligence as a strategic ecosystem capability that enables resilient, adaptive, human-centric, and sustainable manufacturing systems aligned with the emerging objectives of Industry 5.0. ]]>
<![CDATA[Federated Learning for Privacy-Preserving Intelligent Systems ]]>
<![CDATA[Riska Suryani]]>;
<![CDATA[Andri Cahyo Purnomo]]>;
<![CDATA[Arif Budimansyah Purba]]>;
<![CDATA[Leonardus Teguh Handoyo]]>;
<![CDATA[Vindi Tyastutik]]>
<![CDATA[ Technema: Journal of Intelligent Engineering and Computing Vol. 1 No. 2 (2026): : June: Technema: Journal of Intelligent Engineering and Computing ]]>
Publisher : <![CDATA[CV SCRIPTA INTELEKTUAL MANDIRI ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
<![CDATA[The increasing deployment of intelligent systems across healthcare, industrial automation, smart cities, transportation, and edge-computing environments has intensified concerns regarding privacy, data sovereignty, and secure collaborative learning. This study evaluates the effectiveness of Federated Learning (FL) as a privacy-preserving paradigm capable of supporting distributed intelligence without requiring centralized data collection. An empirical experimental design was implemented using a federated architecture consisting of decentralized client nodes, a central aggregation server, the Federated Averaging algorithm, and integrated secure aggregation with differential privacy mechanisms. Experimental evaluation was conducted through repeated validation under heterogeneous client configurations and varying data distributions. The results demonstrate that the proposed framework achieved strong predictive performance, attaining 93.41% accuracy and 95.28% AUC-ROC while maintaining stable convergence under non-identically distributed data conditions. Security evaluation revealed substantial reductions in model inversion, membership inference, and gradient leakage attacks, confirming the effectiveness of the implemented privacy-preserving mechanisms. Scalability analysis further indicated that the framework maintained reliable performance across expanding client populations with acceptable communication overhead and computational efficiency. The findings confirm that Federated Learning provides a practical and scalable foundation for trustworthy intelligent systems by balancing predictive effectiveness, privacy protection, security resilience, and operational feasibility in distributed environments.]]>
<![CDATA[Human-Centered Artificial Intelligence in Intelligent Engineering Systems ]]>
<![CDATA[Rosyid Ridlo Al-Hakim]]>;
<![CDATA[Hasnan Nasrun]]>;
<![CDATA[Fendy Prasetyawan]]>
<![CDATA[ Technema: Journal of Intelligent Engineering and Computing Vol. 1 No. 2 (2026): : June: Technema: Journal of Intelligent Engineering and Computing ]]>
Publisher : <![CDATA[CV SCRIPTA INTELEKTUAL MANDIRI ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
<![CDATA[The increasing integration of artificial intelligence into intelligent engineering systems has created a critical need for approaches that balance computational capability with human autonomy, transparency, accountability, and sustainable operational performance. This study aims to develop an integrative Human-Centered Artificial Intelligence (HCAI) framework capable of supporting trustworthy human–AI collaboration in complex engineering environments. The research employs a non-empirical system-design methodology grounded in design science, systems engineering, socio-technical systems theory, and human-centered AI principles. The proposed framework consists of four interdependent layers comprising human, intelligence, interaction, and governance components that collectively facilitate collaborative decision-making and responsible system operation. Analytical evaluation was conducted through architecture-conformance analysis, requirements-traceability assessment, and scenario-based simulations involving intelligent manufacturing, predictive maintenance, and AI-assisted engineering decision environments. The findings indicate that the framework strengthens human autonomy preservation, explainability capability, collaborative decision efficiency, and governance robustness while maintaining alignment with Industry 5.0 objectives. The study contributes a theoretically integrated architectural model and a reproducible methodological approach for designing intelligent engineering systems that enhance human capabilities, promote trustworthy AI adoption, and support sustainable socio-technical innovation.]]>
