Articles

Access the latest knowledge in applied science, electrical engineering, computer science and information technology, education, and health.

Advanced Search
Filter Icon

Filters article

Years

FAQ Arrow
Single Year
Range Year
0
0

Source Title

FAQ Arrow

Authors

FAQ Arrow

31,456 Article Results

Enhancing academic performance prediction in online learning through hybrid machine learning models

10.11591/ijere.v15i1.33590
Jamal Eddine Rafiq , Zakrani Abdelali , Mohammed Amraouy , Said Nouh
Faced with the rise of online learning platforms, predicting learners’ academic performance has become a major concern to personalize and enhance educational journeys. However, traditional predictive models struggle to effectively integrate emotional and social factors. This article introduces a hybrid predictive model that combines random forests (RF) for selecting the most relevant features and multiple regression (MR) to forecast academic performance. The data is sourced from three online learning platforms and encompasses both implicit traces (learner interactions and behaviors) and explicit traces (demographic characteristics). Following a selection and merging process, the final dataset comprises 1,003,392 records and 42 features, categorized into six types of indicators: cognitive, emotional, social, normative, contextual, and demographic. The results demonstrate that this hybrid model outperforms traditional approaches and other machine learning (ML) techniques in terms of predictive accuracy, achieving an R² of 0.9372 and a root mean square error (RMSE) of 0.1022. The incorporation of explicit and implicit traces helps better capture the intricate interactions among the different data dimensions, significantly enhancing prediction quality. This work represents a notable advancement in the field of academic performance prediction. It also sheds light on challenges associated with the increasing complexity of models, paving the way for future research to develop more generalizable approaches.
Volume: 15
Issue: 1
Page: 436-447
Publish at: 2026-02-01

The role of entrepreneurship education in shaping self-employment intentions: a TPB-based study of Malaysian TVET students

10.11591/ijere.v15i1.35348
Isma Addi Jumbri , Eleeza F. Natasya Khairul Herman , Fauzan Fauzan , Mulyani Karmagatri , Sandy Setiawan , Dian Kurnianingrum
Entrepreneurship education prepares technical and vocational education and training (TVET) students with the competencies and entrepreneurial mindset required for future business endeavors. Guided by the theory of planned behavior (TPB), this study examines how such education shapes students’ self-employment intentions (SEI). A quantitative survey was administered to 300 undergraduates at Universiti Teknikal Malaysia Melaka (UTeM), and the data were analyzed using correlation and regression methods. The results show that entrepreneurial intention is positively linked to attitude toward behavior (r=0.474), entrepreneurship education (r=0.416), subjective norms (r=0.374), and perceived behavioral control (r=0.346), with attitude identified as the most influential predictor, accounting for 22.4% of the variance. These outcomes emphasize the centrality of individual motivation and the enabling role of entrepreneurship education in cultivating entrepreneurial aspirations. The study highlights the strategic importance of embedding stronger entrepreneurship curricula in TVET institutions to enhance self-employment readiness and support national goals for workforce development and economic resilience. The findings resonate with Malaysia’s National Entrepreneurship Policy 2030 and the TVET empowerment agenda, pointing to the value of initiatives such as campus incubators, seed funding, and mentorship in transforming entrepreneurial intention into tangible venture creation.
Volume: 15
Issue: 1
Page: 126-135
Publish at: 2026-02-01

Extended theory of planned behavior: a contextual framework for school mathematics reform

10.11591/ijere.v15i1.36495
Prince Hamid Armah , Robert Benjamin Armah , Dennis Osei Yeboah , Matilda Sarpong Adusei
This study extends the theory of planned behavior (TPB) by testing a structural equation model that incorporates teachers perceived contextual support in explaining implementation of a problem-solving mathematics curriculum reform in Ghana. Using cross-sectional survey data from 368 primary teachers, we measured attitude, subjective norm, perceived behavioral control, intention, self-reported implementation behavior, and contextual support. Confirmatory factor analysis (CFA) supported the measurement model. Structural equation modelling (SEM) showed that attitude (β=.38, p<.001) and perceived behavioral control (β=.29, p<.001) predicted intention, while subjective norm was marginal (β=.12, p=.051). Intention predicted implementation behavior (β=.52, p<.001). Contextual support had a direct effect on behavior (β=.28, p<.001) and strengthened the intention to behavior relationship, which was larger in high support contexts (β=.63) than in low support contexts (β=.30; Δχ²(1)=7.84, p<.01). The model explained 57% of intention and 55% of behavior. Strengthening school resources, leadership support, and professional collaboration is likely to improve mathematics curriculum reform enactment. Policy makers and school leaders should prioritize these contextual supports to help teachers translate mathematics curriculum reform intentions into consistent practice.
Volume: 15
Issue: 1
Page: 725-739
Publish at: 2026-02-01

Assessing the impact of ChatGPT in education: perspectives from Vietnamese educators

10.11591/ijere.v15i1.33610
Nguyen Thi Thu Thuy , Nguyen Thi Kim Lien , Phan Thi Van , Hoang Phuong Lam , Nguyen Le Thanh Truc
ChatGPT is an AI-driven conversational model offering significant benefits for personalized learning, such as adaptability to individual learning styles and increased engagement. While prior research has largely explored student use, limited attention has been given to educators’ perspectives particularly in Vietnam, where access to ChatGPT remains inconsistent. This study examines how perceived information quality (IQ), system quality (SQ), and service quality (SRQ) influence Vietnamese educators’ satisfaction (SF) and intention to use (IU) ChatGPT for academic purposes. Data were gathered through an online survey of 96 educators from five public universities in southern Vietnam and analyzed using partial least squares structural equation modeling (PLS-SEM). Results show that all three quality dimensions significantly affect user SF, but only SQ directly impacts the IU. SF mediates the effects of information and SRQ on usage intentions, emphasizing the importance of user experience in fostering adoption. Policy recommendations include enhancing equitable access to AI tools, promoting AI literacy among educators, and conducting ongoing evaluations to support sustainable implementation. The study highlights how spatial and contextual factors shape AI adoption in education, offering practical insights for integrating ChatGPT and similar tools, especially in under-resourced or geographically remote educational settings.
Volume: 15
Issue: 1
Page: 136-145
Publish at: 2026-02-01

Performance evaluation of serverless cloud-native API deployment: a case study on a mobile health application

10.12928/telkomnika.v24i1.27261
Maulana Bintang; Politeknik Elektronika Negeri Surabaya Irfansyah , Bilal; Politeknik Elektronika Negeri Surabaya Waheed , Idris; Politeknik Elektronika Negeri Surabaya Winarno , Akhmad; Politeknik Elektronika Negeri Surabaya Alimudin
As software applications become increasingly complex, there is a growing need for scalable, flexible, and high-performance backend solutions. Cloud computing-based application programming interfaces (APIs) address these demands by enabling developers to offload resource-intensive tasks to the cloud while eliminating the burden of infrastructure management. This study presents a case study using Obesifix, a mobile health application for real time dietary monitoring and personalized nutrition recommendations. Two deployment models were evaluated: a traditional server-based architecture using Google Compute Engine (GCE) and a serverless approach using Google Cloud Run (GCR). Performance testing was conducted with Apache JMeter under simulated loads of 60, 120, and 180 users across four critical API endpoints (register, login, recommendation, prediction). Results show that GCR consistently achieved 20–30% lower response times and 15–20% higher throughput compared to GCE, while maintaining 0% error rate, lower memory consumption, and more balanced virtual central processing unit (vCPU) utilization. Time to first byte (TTFB) remained below 800 ms across all scenarios, confirming good server responsiveness. These findings highlight the scalability and efficiency benefits of serverless architectures for mobile health applications. Future research should explore asynchronous programming paradigms, autoscaling thresholds, and cost-performance trade-offs, as well as multi-cloud deployments to enhance system resilience and generalizability.
Volume: 24
Issue: 1
Page: 34-48
Publish at: 2026-02-01

The difficulties that beset senior high school students in research writing and presentations

10.11591/ijere.v15i1.35349
Joseph B. Quinto , Girlie C. Jimenez , Illuminada Simbulan Menzi
Writing and presenting research are endeavors that senior high school students have to hurdle in the Philippines. Therefore, this research aimed at unravelling the dimensions of the difficulties faced by senior high school students in their research writing and research presentations through a sequential qualitative mixed-methods design. Primary data was gathered through a qualitative survey of 448 students, while secondary data was obtained via 10 focus groups with 7-10 members representing the different strands in one private senior high school in Baguio City, Philippines. All these participants were selected using combined quota and purposive sampling. Thematic analysis coupled with an iterative procedure was utilized to treat the harvested qualitative data. Findings revealed that the dimensions of the difficulties faced by student-researchers in research writing are health-related, manuscript-related, people-related, and resources-related, while the dimensions in research presentations are inner-directed and outer-directed. This research adds to the current body of knowledge, can contribute to achieving the United Nations sustainable development goal 4.5 and 4.7 on equal education access and global citizenship skills, and recommends research trajectories.
Volume: 15
Issue: 1
Page: 387-398
Publish at: 2026-02-01

RWT-Net: A hybrid ResNet-wavelet-transformer for early detection of left ventricular hypertrophy

10.12928/telkomnika.v24i1.27445
Hoang; The University of Da Nang Huu To Nguyen , Phuong Huu Nghia; San Jacinto College Le , Lam; The University of Da Nang Mai , Nguyen Pham Ho; FPT University Trong
Early detection of left ventricular hypertrophy (LVH), a key predictor of heart failure and stroke, is critical. However, standard 12-lead electrocardiogram (ECG) criteria suffer from low sensitivity. While deep learning shows promise, a research gap exists for models that robustly integrate diverse signal fea tures to improve detection, especially sensitivity. We propose ResNet-wavelet transformer net (RWT-Net), a hybrid architecture that fuses deep morpholog ical features from a ResNet1D with statistical time-frequency features from a wavelet packet transform (WPT) using a transformer encoder. The model was evaluated on the PTB-XL dataset (11,201 recordings) using a stringent, patient level 5-fold cross-validation. RWT-Net achieved a mean area under the curve (AUC)of0.9868andF1-scoreof0.8725. Critically, its wavelet-enhanced stream yielded significantly higher sensitivity compared to a ResNet-transformer base line (0.8964 vs. 0.8716, p=0.0039), better addressing the clinical need to mini mize false negatives. A key limitation is the reliance on ECG-based labels, not an echocardiography gold standard. RWT-Net demonstrates potential as a re liable, automated screening tool to prioritize at-risk patients for further clinical assessment.
Volume: 24
Issue: 1
Page: 142-150
Publish at: 2026-02-01

Optimizing blood cell classification: evaluating feature dimensionality and validation strategies

10.12928/telkomnika.v24i1.27269
Ruaa H. Ali; Northern Technical University Al-Mallah , Marwa Mawfaq; Northern Technical University Mohamedsheet Al-Hatab , Maysaloon; Northern Technical University Abed Qasim
Manual blood cell classification is time consuming and may lead to inconsistent results. This study aims to assist pathologists in diagnosing hematological disorders using machine learning (ML) techniques for automated classification of blood cells in multi-color test images, distinguishing red blood cells (RBCs) and white blood cells (WBCs). Features were extracted using the InceptionV3 network, and several ML models were evaluated for classifying blood cells into eight categories. Two validation strategies: a 66%–34% train–test split and 20-fold cross-validation were applied. The effect of dimensionality reduction through principal component analysis (PCA) was also examined, reducing the feature space from 2,048 to 100 components. Among all models, support vector machine (SVM) achieved highest performance, with 93.4% accuracy and an area under the curve (AUC) of 0.996 without PCA, and 90.1% accuracy with an AUC of 0.991 after PCA. Although PCA slightly reduced accuracy, it improved computational efficiency. Overall, SVM provided the most accurate, stable, and generalizable classification results for automated blood cell analysis.
Volume: 24
Issue: 1
Page: 359-370
Publish at: 2026-02-01

Performance enhancement of embedded object detection via neural hardware acceleration

10.12928/telkomnika.v24i1.27448
Alwin; STMIK Indonesia Mandiri Hartono Limaran , Agung; STMIK Indonesia Mandiri Wicaksono , Patah; STMIK Indonesia Mandiri Herwanto
This paper presents the first benchmarking of you only look once version 11 (YOLO11) on the Rockchip RK3566 neural processing unit (NPU) within the Orange Pi 3B platform. Performance was compared between the quad-core ARM Cortex-A55 CPU and the integrated NPU using the COCO2017 dataset, evaluating latency, energy, and accuracy. NPU acceleration achieved >80% latency reduction and ≈ 94% lower per-inference energy consumption, with speedup of up to 16.7× while maintaining accuracy within 0.03 mean average precision (mAP) of the baseline. Average power remained nearly constant (3.60 W central processing unit (CPU) vs. 3.59 W NPU), indicating that the efficiency gains stem from reduced inference time rather than lower wattage. Limitations included unstable INT8 quantization due to unsupported operators and calibration-range mismatch, as well as minor CPU-side overhead in preprocessing and non-maximum suppression. The findings confirm that the RK3566 NPU delivers substantial efficiency gains without accuracy loss, enabling compact and low-cost platforms to sustain modern object-detection workloads. This demonstrates that affordable NPUs can provide reliable, real time artificial intelligence (AI) inference for embedded vision, internet of things (IoT), and robotics applications.
Volume: 24
Issue: 1
Page: 126-141
Publish at: 2026-02-01

Newchaos function from the composition of DTM and Gauss iterated map for digital image encryption

10.12928/telkomnika.v24i1.27551
Adrianus; Universitas Indonesia Yosia , Tokonyai Tawanda Jonathan; Universitas Indonesia Rabvemhiri , Suryadi; Universitas Indonesia MT
This manuscript introduces a novel chaotic discrete function, formulated through the composition of the dyadic transformation map (DTM) and the Gauss iterated map (GIM), and designated as DTGIM. The National Institute of Science and Technology (NIST) randomness test suite, bifurcation diagrams, and Lyapunov exponents are used to examine the chaotic characteristics of DTGIM. With ini tial condition x0 = 0.12345 and parameters α = −15 and β = 0.3, the func tion shows chaotic behavior in the bifurcation diagram and produces a positive Lyapunov exponent. Strong randomness is further confirmed by NIST tests, which achieve 100% for 32-bit binary sequences and 63.75% for 8-bit binary sequences. Additionally, we compare a number other chaotic discrete functions that also employ the composition method. These findings show that DTGIM is a viable option for applications involving chaos-based cryptography.
Volume: 24
Issue: 1
Page: 228-239
Publish at: 2026-02-01

Performance analysis of a multi-level inverter fed permanent magnet synchronous motor for electric vehicles

10.12928/telkomnika.v24i1.27234
Donepudi; Aditya University Tata Rao , Bhimaraju; Aditya University Pemmanaboidi Srihari Datta , Uma Phanendra; Aditya University Kumar Chaturvedula , Kondala; Aditya University Rao Parasa , Mummidi Parvateeswara; Aditya University Subba Raju
Electric vehicle (EV) drive systems utilizing permanent magnet synchronous motors (PMSMs) often encounter performance limitations due to switching losses, voltage stress, and harmonic distortion. To address these challenges, this paper presents a compact 31-level multilevel inverter (MLI) topology designed to enhance drive efficiency and power quality. The proposed inverter minimizes switching devices and driver circuits, resulting in reduced total harmonic distortion (THD), lower voltage stress, and improved waveform fidelity. Advanced control strategies are employed to further optimize performance. field-oriented control (FOC) ensures precise torque and flux regulation, while direct torque control (DTC) delivers rapid transient response. To mitigate torque ripple and variable switching frequency inherent in conventional DTC, adaptive predictive control (APC) is integrated to refine switching behavior and enhance dynamic stability. Simulation studies conducted in MATLAB/Simulink demonstrate the effectiveness of the proposed system, revealing significant improvements in torque smoothness, reduced THD (0.85%) and elevated efficiency under variable load conditions. This integrated solution offers a practical and scalable approach for next-generation EVs, contributing to greater reliability, energy utilization, and overall system performance.
Volume: 24
Issue: 1
Page: 302-312
Publish at: 2026-02-01

Secure hybrid power-frequency multiple access in satellite terrestrial communication systems: a performance study

10.12928/telkomnika.v24i1.26892
Huu; Industrial University of Ho Chi Minh City Q. Tran , Viet-Thanh; Industrial University of Ho Chi Minh City Pham
This paper investigates a secure hybrid power–frequency multiple access (PFMA) framework for satellite–terrestrial communications. By integrating power- and frequency-domain multiplexing, PFMA achieves approximately 4 dB lower transmit signal-to-noise ratio (SNR) than non-orthogonal multiple access (NOMA) for the same connection outage probability (COP) at SNR > 0 dB, and it reduces the COP by up to 30% at low-to-medium SNRs. It further decreases the intercept probability (IP) by 20–25% at PS = 10 dBm. Closed-form COP and IP expressions are derived under shadowed-Rician fad ing with both internal and external eavesdroppers and validated via Monte Carlo simulations. Parameter analysis indicates that PFMA’s SNR gain can either ex tend coverage by 60% or save 37% energy, providing design guidelines for 6G, satellite IoT, and emergency communication systems. The single-cell assump tion points to future work on multi-cell and mobility scenarios.
Volume: 24
Issue: 1
Page: 14-21
Publish at: 2026-02-01

Deep learning-based power amplifier linearization in OFDM systems with unknown channel state information

10.12928/telkomnika.v24i1.27236
Meryem Mamia; University of Tlemcen Benosman , Mohammed Yassine; University of Tlemcen Bendimerad , Fethi Tarik; University of Tlemcen Bendimerad
This paper presents an end-to-end deep learning-based approach for orthogonal frequency-division multiplexing (OFDM) communication systems impaired by nonlinear power amplifiers (PAs) and channel fading. The PA nonlinearity is modeled using the modified Rapp model, and simulations are performed on a 64-subcarrier OFDM system with a cyclic prefix (CP) of 8 and 16-quadrature amplitude modulation (16-QAM). The proposed autoencoder-based OFDM–PA (AE-OFDM-PA) system jointly optimizes the transmitter and receiver through end-to-end learning, enabling simultaneous compensation of both PA nonlinearities and channel distortions without requiring explicit channel state information (CSI) estimation. Instead, the model leverages embedded pilot sequences to learn the implicit CSI representation directly from data, allowing the receiver to correct amplitude and phase distortions adaptively. Simulation results demonstrate that AE-OFDM-PA significantly outperforms conventional OFDM and OFDM-PA systems, achieving over 70× block error rate (BLER) improvement compared with the uncompensated OFDM-PA system at an input back-off (IBO) of 3 dB. Furthermore, the proposed method achieves approximately 11.5 dB adjacent channel leakage ratio (ACLR) improvement over the classical memory polynomial digital predistortion (DPD) technique, while slightly reducing the peak-to-average power ratio (PAPR). Overall, AE-OFDM-PA provides a robust, spectrally efficient, and low-complexity solution for nonlinear and fading environments with unknown or varying CSI.
Volume: 24
Issue: 1
Page: 1-13
Publish at: 2026-02-01

Information technology value engineering through partial adjustment valuation theory

10.12928/telkomnika.v24i1.27478
Lukman; Telkom University Abdurrahman , Candiwan; Telkom University Candiwan
The paper proposes a systems management approach that utilizes information technology (IT) treatment as a framework to help firms enhance future performance by optimising key parameters. The method certifies a valuation approach that enables businesses to better manage their IT infrastructure and improve performance. A case study of A case study of PT Telekomunikasi Indonesia (Telkom) and PT XL Axiata (XL) (2004–2018) shows the method’s effectiveness. Once the IT value is identified, specific parameters can be engineered to improve performance without changing other variables. The approach uses a partial adjustment valuation model, enabling performance gains at lower costs. The results show significant improvements in both firms’ performance values and ratios compared to their originals. This supports adopting a cost leadership strategy, making IT based businesses more efficient, cost-effective, and better performing across financial, business, and strategic dimensions.
Volume: 24
Issue: 1
Page: 111-125
Publish at: 2026-02-01

Advancements in physical layer key generation: a review on channel reciprocity and IoT security techniques

10.12928/telkomnika.v24i1.27340
Syed Shafaq; Southeast University Ali Shah , Ajab; University of Science and Technology Bannu Noor , Ruiyue; Changchun University of Science and Technology Liang , Rahmat Ullah; FAST National University of Computer and Emerging Science Zadran
With the burgeoning internet of things (IoT), securing communication becomes paramount. Traditional cryptography does not meet computational needs and brute-force attacks. This review explores the state-of-the-art physical layer secret key generation (PLKG) that takes advantage of the inherent reciprocity and randomness of wireless channels. We investigate cutting-edge techniques such as feature extraction networks, domain adversarial training, and deep learning-based approaches, evaluating their effects on the security and efficiency of key generation. In addition to these methods, the review addresses real-world challenges such as multi-user scenarios, reconciliation overhead, and inconsistent channel measurement. We believe that improved key generation rates and security can be achieved through the use of millimeter wave technology and full-duplex communication. To strengthen the robustness of key generation, the paper concludes by suggesting future directions, such as incorporating more random sources, such as physiological signals and sensor data. This comprehensive overview offers deep insights into the state-of-the-art and paves the way for reliable communication in ever more complicated IoT settings.
Volume: 24
Issue: 1
Page: 196-205
Publish at: 2026-02-01
1 2 3
...
2097 2098
Show 151 of 2098

Discover Our Library

Embark on a journey through our expansive collection of articles and let curiosity lead your path to innovation.

Explore Now
Library 3D Ilustration