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 acoustic environment classification for hearingimpaired individuals using hybrid CNN and RFE

10.11591/ijeecs.v39.i2.pp906-913
Sunilkumar M. Hattaraki , Shankarayya G. Kambalimath
Individuals who are deaf or hard of hearing experience considerable difficulties in distinguishing sounds in various acoustic environments, which affects their communication ability and overall quality of life. Existing auditory assistive technologies currently face challenges with real-time classification and adaptation to changing noise conditions, underscoring the need for more reliable and accurate classification models. This research bridges the existing gap by creating a hybrid classification framework that integrates convolutional neural networks (CNN) and random forest ensemble (RFE) to enhance the accuracy of environmental sound classification. The study utilizes Mel-frequency cepstral coefficients (MFCCs) for feature extraction and principal component analysis (PCA) for dimensionality reduction, thus facilitating the efficient processing of real-world audio data. The proposed methodology improves classification accuracy across various environmental conditions. Experimental evaluations demonstrate superior performance, achieving a training accuracy of 94.93% and a testing accuracy of 93.41%, thereby exceeding conventional machine learning methods. By overcoming limitations in existing models, this research contributes to the development of adaptive hearing assistance systems with enhanced noise classification capabilities. The results have significant implications for the development of smart hearing aids, real-time noise classification, and auditory scene analysis. Ultimately, this research enhances assistive hearing technologies, promoting greater accessibility, communication, and inclusion for hearing-impaired individuals, thus contributing positively to society.
Volume: 39
Issue: 2
Page: 906-913
Publish at: 2025-08-01

Fuel consumption prediction of civil air crafts using deep learning: a comparative study

10.12928/telkomnika.v23i4.26449
Quoc Hung; University of Economics Ho Chi Minh City (UEH) Nguyen , Hoang Lan; University of Economics Ho Chi Minh City (UEH) Nguyen
Accurate fuel consumption prediction is critical for minimizing the adverse impact of fuel emissions on the environment, conserving fuel, and reducing flight costs. Additionally, precise fuel forecasting enhances trajectory prediction and supports effective air traffic management. This study evaluates the predictive performance of two deep learning techniques in predicting the fuel consumption of a civil aircraft belonging to Airbus A320NEO. Based on the analysis, the findings show that the deep neural network (DNN) model has better score of indicators and than the recurrent neural network (RNN) including mean absolute error (MAE), mean squared error (MSE), root mean squared error (RMSE) and R-squared (R2). By integrating an automated feature selection approach with an optimized deep learning framework, this research contributes to the development of a robust and efficient predictive system for fuel consumption. The findings have practical implications for improving fuel management strategies in aviation, leading to cost savings and reduced emissions. One limitation of this study is its reliance on specific environmental variables, which may limit the model’s generalizability across different flight conditions, aircraft types, and operational scenarios.
Volume: 23
Issue: 4
Page: 1000-1009
Publish at: 2025-08-01

A convolution neural network model for knee osteoporosis classification using X-ray images

10.12928/telkomnika.v23i4.26778
Omar; University of Mosul Khalid M. Ali , Abeer; University of Mosul K. Ibrahim , Bilal; University of Mosul R. Altamer
Bone structure deterioration along with low levels of bone density are the hallmarks of knee osteoporosis (KOP). The conventional approach for detecting osteoporosis is accomplished using a knee radiograph, but it requires specialized knowledge. Nevertheless, X-rays can be difficult to interpret due to their large volume and minor fluctuations. In the past few decades, deep learning algorithms have minimized misinterpretation and modified medical diagnosis. In particular, algorithms based on convolutional neural networks (CNNs) have been used to speed up the procedure of diagnosis because of their innate capacity to extract significant features that often are challenging to spot by hand. A robust CNN model was proposed in this paper for KOP classification which uses a train and test approach to recognize healthy, osteopenia-predicted, and osteoporosis knee cases using 1947 X-ray images. The proposed model was designed using Jupyter Notebook and is in Python. To verify the efficiency of the model, some factors were calculated such as accuracy, precision, recall, and f1-score. In comparison with other similar systems, the results obtained showed that the accuracy of the proposed system reached 90.25%.
Volume: 23
Issue: 4
Page: 1010-1019
Publish at: 2025-08-01

Business intelligence system model to measure the performance of lecturers’ scientific publications

10.12928/telkomnika.v23i4.26221
Miwan Kurniawan; Universitas Bina Sarana Informatika Hidayat , Dedy; Universitas Trisakti Sugiarto , Rina; Universitas Trisakti Fitriana , Yun-Chia; Yuan Ze University Liang
Scientific publication data is sourced from the SINTA website to measure the performance of journals, institutions, and researchers in Indonesia. Accessing and analyzing data for institutional needs is restricted, and lecturer development patterns based on lecturer characteristics remain untapped. The study aims to analyze and design business intelligence system models to measure the performance of scientific publications using dimensional models, clustering, on-line analytical processing (OLAP), and prototyping. Research methods are carried out through data and information needs analysis, data warehouse design, data mining and OLAP application, business intelligence system development, and system evaluation. The resulting dimensional models are the researcher index model, the researcher score model, the publication article model, and the research subject model. Measurements of data size and processing time show that the star schema has data of 336 KB and a processing time of 0.00554 seconds, is the best model compared to the snowflake’s schema, which has data of 368 KB and a processing time of 0.00611 seconds. Davies–Bouldin Index (DBI) measurements show the best clustering performance is the X-means algorithm with K as many as 5 clusters (Kmin=3, Kmax=5) and a DBI value of 0.537040.
Volume: 23
Issue: 4
Page: 954-964
Publish at: 2025-08-01

Artificial neural network based load flow analysis of radial distribution system in Kurdistan region

10.11591/ijeecs.v39.i2.pp761-773
Warda Hussein Ali , Dana O. Qader , Mohamed A. Hussein
Today electric energy is the most commonly used source in the world. Power flow (load flow) analysis is conciderd as the backbone of any power system analysis and design; they have a great necessity for operating systems, future planning, fault analysis, and contingency analysis. For better utilization of electrical power, off-line modeling and simulation of power systems using powerful software are essential and significant task especially in developing countries and regions. Therefore, this paper performs a comparison study of conventional and non-conventional load flow techniques for a 24-Bus radial distribution system in the governorate of Sulaymaniyah. The conventional power flow techniques include the Newton-Raphson (NR), and Gauss-Seidel (GS) techniques, while the nonconventional load flow technique utilizes the artificial neural network (ANN). Modeling, simulation, and analysis of the 24-Bus feeder are performed using MATPOWER simulation tool. The MATPOWER and neural network techniques are implemented independently, and it has been proved that ANN model efficiently estimated the power flow analysis for the system mentioned above, the high regression values of nearly 0.999 indicates that the ANN model can be used as an efficient tool to perform power flow analysis.
Volume: 39
Issue: 2
Page: 761-773
Publish at: 2025-08-01

An improved approximate parallel prefix adder for high performance computing applications: a comparative analysis

10.11591/ijict.v14i2.pp382-392
Vamsidhar Anagani , Kasi Geethanjali , Anusha Gorantla , Annamreddy Devi
Binary adders are fundamental in digital circuit designs, including digital signal processors and microprocessor data path units. Consequently, significant research has focused on improving adders’ power-delay efficiency. The carry tree adder (CTA) is alternatively referred to as the parallel prefix adder (PPA), is among the fastest adders, achieving superior performance in very large scale integrated (VLSI) implementations through efficient concurrent carry generation and propagation. This study introduces approximate PPAs (AxPPAs) by applying approximations in prefix operators (POs). Four types of AxPPAs approximate kogge-stone, approximate brent-kung, approximate ladner fischer, and approximate sparse kogge-stone-were designed and implemented on FPGA with bit widths up to 64-bit. Delay measurements from static timing analysis using Xilinx ISE design suite version 14.7 indicate that AxPPAs exhibit better latency performance than traditional PPAs. The AxPPA sparse kogge-stone, in particular, demonstrated superior area and speed performance, achieving a delay of 2.501ns for a 16-bit addition.
Volume: 14
Issue: 2
Page: 382-392
Publish at: 2025-08-01

Smart hybrid power management system in electric vehicle for efficient resource utilization with ANN

10.11591/ijict.v14i2.pp488-496
Vinoth Kumar Ponnusamy , Gunapriya Devarajan , Gomathi Easwaram , Geetha Murugesan , Rathinam Marimuthu Sekar , R. Delshi Howsalya Devi
The novel hybrid power system integrating energy storage, electric vehicle (EV) charging infrastructure and renewable energy sources improve sustainability and resilience. This work proposes a power management system controlled by artificial intelligence for EV charging infrastructure. The battery’s state of charge (SoC) is continuously monitored by artificial neural network (ANN) algorithm improves the health of the battery and efficient operation of the system. The buck boost DC-DC converter performs dynamic switching mechanism, which adjusts to changing solar conditions and energy demands, guarantees a steady power supply. Depending on the situation, the ANN algorithm used in the battery’s SoC control mechanism decides whether to priorities the EV charging or the inverter to supply the grid. The work is evaluated with the MATLAB simulation for different conditions and results are compared with different controllers such as PI, PID, and ANN. The experiment performed uses internet of things (IoT) for transferring the data from the EV motor, acts as an input for the controller to perform the novel hybrid power management operation with cloud technology. The experimental prototype evaluates the results connected to the photovoltaic (PV) system and battery management system (BMS) which lowers reliance on non-renewable resources, increases overall energy efficiency, and ensures a steady supply of power under a various condition.
Volume: 14
Issue: 2
Page: 488-496
Publish at: 2025-08-01

An implementation of GAN analysis for criminal face identification system

10.11591/ijeecs.v39.i2.pp963-972
Ayesha Sarosh , Govindu Komali , Vishnu Vardhan Battu , Laxmaiah Kocharla , Eswaree Devi Kopparavuri , Ooruchintala Obulesu , Praveen Mande , Amanulla Mohammad
In recent times, the criminal activities are growing at an exponential rate. For the prevention of crime, one of the main issues that are before the police are accurate identification of criminals and on the other hand the availability of police officers are not adequate. The most tedious task is tracking the suspect once a crime was committed. Over the years, several technical solutions have been presented to detect the criminals however most of them were not effective. One of the most significant characteristics for the identification of a person is face. Even identical twins have their own unique faces. Face identification is a challenging topic in computer vision because the human face is a dynamic entity with a high degree of visual variation. In this area, identification accuracy and speed are significant challenges. Hence to solve these issues, an implementation of generative adversarial network (GAN) analysis for criminal face identification system is presented. GAN is used for the identification of criminals. Recall, precision, accuracy, and F1-score are used to assess the performance of the presented technique. Compared to previous models, this model will achieve better performance for criminal face detection.
Volume: 39
Issue: 2
Page: 963-972
Publish at: 2025-08-01

Neural control of DVR for wind turbine grid fault mitigation with PIL validation

10.11591/ijeecs.v39.i2.pp797-806
Kaoutar Dahmane , Belkasem Imodane , Said Mailal , Brahim Bouachrine , Mohamed Ajaamoum , Mhand Oubella
Power quality issues that include voltage sag and swell challenge grid stability, not least for renewable energy systems such as wind turbines (WTs). Occurrence of these voltage disturbances impacts severely the performance of WT systems, compromising their fault ride-through (FRT) capabilities. This work investigates the application of an artificial neural network (ANN) as a controller mechanism for a dynamic voltage restorer, aimed at improving the FRT capabilities of a WT equipped with a permanent magnet synchronous generator. The approach includes employing series compensation to maintain the terminal voltage of the WT during fault conditions. This is performed by injecting voltage at the interface where the system connects to the grid, thus stabilizing the terminal voltage within the wind energy system. The control of the dynamic voltage restorer (DVR) is fundamental to improve the FRT capability. An ANN approach, as control technique is applied to drive the DVR. Training data used for ANN are obtained from a proportional-integral controller, and the proposed system is comprehensively modeled with MATLAB/Simulink. The proposed method demonstrates effective voltage restoration, under two fault scenarios: voltage sag and swell. Besides, the processor in-the-loop (PIL) test proves that the suggested control is practically implementable.
Volume: 39
Issue: 2
Page: 797-806
Publish at: 2025-08-01

Comparison of robust machine learning algorithms on outliers and imbalanced spam data

10.11591/ijeecs.v39.i2.pp1130-1144
Dodo Zaenal Abidin , Jasmir Jasmir , Errisya Rasywir , Agus Siswanto
Effective spam detection is essential for data security, user experience, and organizational trust. However, outliers and class imbalance can impact machine learning models for spam classification. Previous studies focused on feature selection and ensemble learning but have not explicitly examined their combined effects. This study evaluates the performance of random forest (RF), gradient boosting (GB), and extreme gradient boosting (XGBoost) under four experimental scenarios: (i) without synthetic minority over-sampling technique (SMOTE) and outliers, (ii) without SMOTE but with outliers, (iii) with SMOTE and without outliers, and (iv) with SMOTE and with outliers. Results show that XGBoost achieves the highest accuracy (96%), an area under the curve-receiver operating characteristic (AUCROC) of 0.9928, and the fastest computation time (0.6184 seconds) under the SMOTE and outlier-free scenario. Additionally, RF attained an AUCROC of 0.9920, while GB achieved 0.9876 but required more processing time. These findings emphasize the need to address class imbalance and outliers in spam detection models. This study contributes to developing more robust spam filtering techniques and provides a benchmark for future improvements. By systematically evaluating these factors, it lays a foundation for designing more effective spam detection frameworks adaptable to real-world imbalanced and noisy data conditions.
Volume: 39
Issue: 2
Page: 1130-1144
Publish at: 2025-08-01

Study of design thinking and software engineering integration in education and training

10.11591/ijeecs.v39.i2.pp1384-1398
Muhammad Ihsan Zul , Suhaila Mohd. Yasin , Dadang Syarif Sihabudin Sahid
Integrating design thinking (DT) with software engineering (SE) is widely applied in industry, serving as a reference for SE in education and training. The industry has various integration models, but researchers and educators mainly adapt them for education. A clear understanding of DT-SE integration models is essential to figuring out their implementation. This study examines existing DT-SE integration models, challenges, and integration methods using Kitchenham’s framework in education and training. The paper was collected from ScienceDirect, IEEEXplore, Scopus, ACM, SpringerLink, and Google Scholar, yielding 593 initial publications, with 43 selected for in-depth analysis. Findings indicate that the d.school model is the most widely adopted DT model. Key challenges include team dynamics, process management, complexity, and cultural factors. DT is integrated into requirements engineering (RE) due to its user-centered nature, though only two studies explicitly describe DT-SE integration models, both applied early in SE processes. These findings suggest educational practices align with industry trends in model adoption and integration focus. Educators and practitioners can use these insights to design or adapt integration models suitable for education and training by shaping curricula that emphasize user-centered design, collaboration, and the extension of DT practices beyond RE-strengthening its impact for education and training.
Volume: 39
Issue: 2
Page: 1384-1398
Publish at: 2025-08-01

Designing an automated matching model to enhance recruitment process

10.11591/ijeecs.v39.i2.pp1081-1091
Sahar Idwan , Ebaa Fayyoumi , Haneen Hijazi , Izzeddin Matar
Detecting qualified candidates for a vacant position is a difficult task, especially when there are numerous applicants. This delays team development in finding the appropriate individual at the right moment. Adopting a well-structured selection process will create opportunities for new aspects and ideas. In this paper, the matching job applicant (MJA) model is developed to assist all parties, the employers and the employees simultaneously by providing a fair, transparent unbiased solution constructed by using a mathematical machine. This provides a clear justification in the decision-making process in addition to advising the applicants with the most suitable positions that fits their qualifications.
Volume: 39
Issue: 2
Page: 1081-1091
Publish at: 2025-08-01

Real-time recognition of Indonesian sign language SIBI using CNN-SVM model combination

10.11591/ijeecs.v39.i2.pp1198-1210
Satriadi Putra Santika , Stefanus Benhard , Yulyani Arifin , Andry Chowanda
Real-time Sistem Isyarat Bahasa Indonesia (SIBI) sign language recognition plays a crucial role in improving accessibility for individuals with hearing and speech impairments. Despite advancements in SIBI recognition research, challenges remain in ensuring model stability and accuracy in realtime settings, particularly in handling gesture variations and classification inconsistencies. This study addresses these challenges by developing a convolutional neural network-support vector machine (CNN-SVM) combination model, integrating MediaPipe for hand coordinate extraction, CNN for feature extraction, and SVM for classification. To improve generalization and prevent overfitting, data augmentation is applied to expand the dataset. The model's performance is further enhanced through hyperparameter optimization (HPO) and post-processing techniques such as multi-window majority voting (MWMV) and SymSpell. Experimental results show that the CNN-SVM model trained on augmented data with HPO achieves 91% testing accuracy, outperforming both standalone CNN and SVM models. Furthermore, MWMV improves recognition stability, while SymSpell enhances spelling errors, ensuring more meaningful outputs. The system is integrated with OpenCV for real-time recognition, but current deployment remains limited to local execution. Future work will focus on developing lightweight models for web-based and mobile applications, making the system more accessible and scalable.
Volume: 39
Issue: 2
Page: 1198-1210
Publish at: 2025-08-01

Deep learning algorithms for breast cancer detection from ultrasound scans

10.11591/ijict.v14i2.pp427-437
Lawysen Lawysen , Gede Putra Kusuma
Breast cancer is a highly dangerous disease and the leading cause of cancer related deaths among women. Early detection of breast cancer is considered quite challenging but can offer significant benefits, as various treatment interventions can be initiated earlier. The focus of this research is to develop a model to detect breast cancer based on ultrasound results using deep learning algorithms. In the initial stages, several preprocessing processes, including image transformation and image augmentation were performed. Two types of models were developed: utilizing mask files and without using mask files. Two types of models were developed using four deep learning algorithms: residual network (ResNet)-50, VGG16, vision transformer (ViT), and data-efficient image transformer (DeiT). Various algorithms, such as optimization algorithms, loss functions, and hyperparameter tuning algorithms, were employed during the model training process. Accuracy used as the performance metric to measure the model’s effectiveness. The model developed with ResNet-50 became the best model, achieving an accuracy of 94% for the model using mask files. In comparison, the model developed with ResNet-50 and DeiT became the best model for the model without mask files, with an accuracy of 80%. Therefore, it can be concluded that using mask files is crucial for producing the best-performing model.
Volume: 14
Issue: 2
Page: 427-437
Publish at: 2025-08-01

A novel-shaped THz MIMO antenna with high bandwidth for advanced 6G wireless application

10.12928/telkomnika.v23i4.26559
Kamal Hossain; Daffodil International University Nahin , Jamal; Daffodil International University Hossain Nirob , Md. Ashraful; Daffodil International University Haque , Narinderjit Singh; INTI International University Sawaran Singh , Redwan Al Mahmud; Daffodil International University Bin Asad Ananta , Md. Kawsar; Daffodil International University Ahmed , Md. Sharif; Daffodil International University Ahammed , Liton; Pabna University of Science and Technology Chandra Paul
This article presents an industrial and innovation highly efficient drone shaped slotted graphene-based multiple input multiple output (MIMO) antenna with improved isolation, designed for high-speed short-range communication, video rate imaging, medical imaging, and explosive detection in the THz band. The proposed antenna is constructed on an 88×244 μm2 polyimide substrate. Key performance parameters such as reflection coefficient, gain, directivity, radiation pattern, and antenna efficiency are evaluated at the resonating frequencies of 1.7 THz, 3.35 THz, and 5.31 THz, covering a wide bandwidth of 4.88 THz with a reflection coefficient of less than -10 dB. The antenna achieves a maximum gain of 13.92 dB and a radiation efficiency of 95.77% within the resonating band. The MIMO design parameters include an envelope correlation coefficient (ECC) of 0.00015, a diversity gain (DG) of 9.9992, and an isolation of less than -31.55 dB between its elements across the entire bandwidth. The outcomes from CST simulations were verified by designing and simulating a similar resistance-inductance-capacitance (RLC) circuit in advanced design system (ADS), with both simulators producing comparable reflection coefficients. These features underscore the potential of the proposed antenna, utilizing simulations and an equivalent RLC circuit model, as a robust candidate for THz band applications in 6G wireless communication.
Volume: 23
Issue: 4
Page: 847-856
Publish at: 2025-08-01
1 2 3
...
2097 2098
Show 249 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