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

Optimal control of the dynamics of nonlinear oscillating systems using synergetic principles of self-organization

10.11591/ijece.v14i6.pp6271-6278
Siddikov Isomiddin Xakimovich , Umurzakova Dilnoza Maxamadjanovna
This paper analyses the evolution of nonlinear oscillation control methods and presents an innovative approach known as analytical design of aggregated oscillation controllers (ADACO). This method is based on the principles of synergetic control theory and focuses on the integration of self-organization and control processes to synthesize energy-efficient control laws for nonlinear oscillating systems. The authors elaborate on the theoretical foundations of ADACO, which extends the previous analytical design of aggregated controllers (ADAC) method by incorporating energy invariants and integrals of motion into the synthesis of control laws. This approach demonstrates significant advantages over traditional methods, offering a versatile framework for the design of energy-efficient control systems for a wide range of nonlinear oscillating systems in various fields such as aerospace, robotics, vibromechanical systems, and objects with chaotic dynamics. The aim of the paper is to establish a unified approach to the control of nonlinear oscillations, solving both the problems of generation of stable oscillations and suppression of unwanted perturbations. The application of synergetic control principles in the framework of ADACO opens prospects for further development of nonlinear control theory.
Volume: 14
Issue: 6
Page: 6271-6278
Publish at: 2024-12-01

Framework for detecting and resisting cyberattacks on cyber-physical systems in internet of things

10.11591/ijece.v14i6.pp7169-7177
Jyoti Metan , Mahantesh Mathapati , Prasad Adaguru Yogegowda , Kurilinga Sannalingappa Ananda Kumar
Cyber-physical system (CPS) is an integral part of an internet of things (IoT) with established wide spread applications. An increasing concern towards being highly vulnerable to various forms of dynamic cyber-attacks has been increasingly evolving. A review of existing research methodology showcases complex solutions that can offer sub-optimal security strength when exposed to dynamic cyber-attack forms while increasing the computational burden. Therefore, this manuscript presents a novel yet simplified computational framework capable of determining and resisting critical anomalies within internet-of-cyber physical systems (IoCPS). The presented scheme contributes towards preprocessing following a distinct oversampling method targeting balancing the data. An ensemble machine learning model using a discrete variant of AdaBoost and neural decision tree (NDT) has been implemented to optimize the learning process and improve the threat detection efficiency. The comparative outcome of the proposed study showcases that it offers approximately 7.2% increased threat detection accuracy and approximately 68% reduced response time compared to frequently adopted learning mechanisms towards threat detection over an IoT environment.
Volume: 14
Issue: 6
Page: 7169-7177
Publish at: 2024-12-01

Traffic signs detection and prohibitor signs recognition in Morocco road scene

10.11591/ijece.v14i6.pp6313-6321
Imane Taouqi , Abdessamad Klilou , Kebir Chaji , Assia Arsalane
Traffic sign detection is a crucial aspect of advanced driver assistance systems (ADAS) for academic research and the automotive industry. seeing that accurate and timely detection of traffic signs (TS) is essential for ensuring the safety of driving. However, TS detection methods encounter challenges like slow detection speed and a lack of robustness in complex environments. This paper suggests addressing these limitations by proposing the use of the you only look one version 7 (YOLOv7) network to detect and recognize TS in road scenes. Furthermore, the k-means++ algorithm is used to acquire anchor boxes. Additionally, a tiny version of YOLOv7 is used to take advantage of its real-time and low model size, which are required for real-time hardware implementation. So, we conducted an experiment using our proprietary Morocco dataset. According to the experimental results, YOLOv7 achieves 85% in terms of mean average precision (mAP) at 0.5 for all classes. And YOLOv7-tiny obtains 90% in the same term. Afterward, a recognition system for the prohibitive class using the convolutional neural network (CNN) is trained and integrated inside the YOLOv7 algorithm; its model achieves an accuracy of 99%, which leads to a good specification of the prohibitive sign meaning.
Volume: 14
Issue: 6
Page: 6313-6321
Publish at: 2024-12-01

ConciseCarNet: convolutional neural network for parking space classification

10.11591/ijai.v13.i4.pp4158-4168
Marwan Ramli , Sayuti Rahman , Rahmad Bayu Syah
The car is a mode of transportation that brings numerous benefits to the community. As a result, the growth of vehicles is increasing, which has a negative impact. Some of the negative impacts include noise, air pollution, traffic congestion, and the need for parking spaces. Drivers that drive around looking for parking places increase the negative impact as well as boredom and even worry for the driver. Therefore, the driver needs this information on the availability of parking spaces. A convolutional neural network (CNN) using a camera is one of the best methods that can be used to solve this problem. We built a more efficient CNN architecture for classifying parking spaces, which was named ConciseCarNet. ConciseCarNet uses 33 and 11 convolution filters, which cause fewer parameters than previous architectures. ConciseCarNet has two branches, each with a different branch structure. This branch is designed to generate additional feature variations, which will help improve the accuracy. Based on testing, the accuracy of ConciseCarNet2x outperforms the accuracy of mAlexnet, Carnet, EfficientParkingNet, and you look once (YOLO)+MobilNet architectures, which is 99.37%. ConciseCarNet has fewer parameters, file sizes, and floating point operations (FLOPs) compared to other architectures.
Volume: 13
Issue: 4
Page: 4158-4168
Publish at: 2024-12-01

Braille code classifications tool based on computer vision for visual impaired

10.11591/ijece.v14i6.pp6992-7000
Hany M. Sadak , Ashraf A. M. Khalaf , Aziza I. Hussein , Gerges Mansour Salama
Blind and visually impaired people (VIP) face many challenges in writing as they usually use traditional tools such as Slate and Stylus or expensive typewriters as Perkins Brailler, often causing accessibility and affordability issues. This article introduces a novel portable, cost-effective device that helps VIP how to write by utilizing a deep-learning model to detect a Braille cell. Using deep learning instead of electrical circuits can reduce costs and enable a mobile app to act as a virtual teacher for blind users. The app could suggest sentences for the user to write and check their work, providing an independent learning platform. This feature is difficult to implement when using electronic circuits. A portable device generates Braille character cells using light- emitting diode (LED) arrays instead of Braille holes. A smartphone camera captures the image, which is then processed by a deep learning model to detect the Braille and convert it to English text. This article provides a new dataset for custom-Braille character cells. Moreover, applying a transfer learning technique on the mobile network version 2 (MobileNetv2) model offers a basis for the development of a comprehensive mobile application. The accuracy based on the model reached 97%.
Volume: 14
Issue: 6
Page: 6992-7000
Publish at: 2024-12-01

Optimal control strategies based on extended Kalman filter in mathematical models of COVID-19

10.11591/ijece.v14i6.pp6300-6312
Dewi Suhika , Roberd Saragih , Dewi Handayani , Mochamad Apri
The Omicron variant of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus is an extremely contagious variant that has garnered global attention due to its potential for rapid spread and its impact on the effectiveness of vaccines and non-pharmacological measures. In this paper, we investigate mathematical models involving vaccinated individuals and control functions to analyze how the spread of coronavirus disease 2019 (COVID-19) infection evolves over time. In the process of constructing a mathematical model for COVID-19, there are many parameters whose values are not yet known with certainty. Therefore, the extended Kalman filter method is used as a tool to estimate these parameters in an effort to better understand the dynamics of the spread and evolution of this disease. This method helps align the mathematical model with existing empirical data, allowing us to make more accurate predictions about the course of the COVID-19 pandemic and plan more precise actions to address the situation. Furthermore, an optimal control design is applied to reduce the number of infected individuals by implementing seven strategies involving a combination of health education, vaccination, and isolation controls. The simulation results we conducted indicate that the use of optimal control strategies can lead to a significant decrease in the number of individuals infected with COVID-19.
Volume: 14
Issue: 6
Page: 6300-6312
Publish at: 2024-12-01

Support vector machine method for classifying severity of Alzheimer's based on hippocampus object using magnetic resonance imaging modalities

10.11591/ijece.v14i6.pp6322-6331
Retno Supriyanti , Arif Pujo Riyanto , Yogi Ramadhani , Muhammad Syaiful Aliim , Muhammad Irham Akbar , Haris Budi Widodo , Muhammad Alqaaf
Alzheimer's disease is a degenerative brain condition that causes progressive decline in several aspects. Starting from memory, cognitive or thinking abilities, speaking abilities, and behavior. Currently, Alzheimer's diagnosis uses some methods, such as blood tests, scanning with computerized tomography scan (CT scan), or magnetic resonance imaging (MRI). As a reference for determining the level of severity, doctors usually use clinical dementia rating (CDR). CDR is a numerical scale used to measure the severity of dementia symptoms. The doctor will manually compare the patient's condition with those stated on the CDR. This condition will take quite a long time, and sometimes human error will occur. As technology and science develop, doctors can assist in manually detecting Alzheimer's using classification algorithms. Many methods can be used to classify, including the CDR support vector machine (SVM) method. Unfortunately, this method is usually only used to classify two classes. This technology allows the classification process to be carried out automatically and quickly. On the other hand, when using CDR to classify Alzheimer's severity, there are several scales, not just two classes. So, in this research, we modified the use of SVM to classify three levels of severity, namely scale 0 for normal, scale 1 for mild conditions, and scale 2 for moderate conditions. The experiments we carried out provided an accuracy of 90.9%.
Volume: 14
Issue: 6
Page: 6322-6331
Publish at: 2024-12-01

Faculty Assistant Bot-automation of administrative activities using robotic process automation

10.11591/ijece.v14i6.pp6797-6806
Mamidyala Durga Prasad , Balusu Nandini
In this paper, a process workflow for Bot is created using robotic process automation (RPA), associated with artificial intelligence that is used to stream line the administrative tasks and alleviate stress levels of faculty in handling administrative tasks while teaching in higher education. These activities are must for National Academic Audit Council (NAAC) accreditation and All India Survey on Higher Education (AISHE) surveys, which strive to bring quality in teaching higher education by shaping educational policies. Therefore, ensuring the accuracy of this data is paramount to avoid misleading decisions. The Bot automatically gathers student results from the website and saves them into individual files, eliminating the need for human intervention. It is trained to find the related file of student and update his results of upcoming semesters or backlogs. The Bot efficiently manages folders during file saving to enhance retrieval. Additionally, it maintains pertinent student details such as community, caste, and religion, beneficial for educational policy surveys aiming for improved quality. Moreover, it generates and updates reports post each process execution, ensuring data integrity, and can be trained for statistical analysis to predict student outcomes. The UiPath tool is used in the design and testing of the developed Bot
Volume: 14
Issue: 6
Page: 6797-6806
Publish at: 2024-12-01

A systematic review of multi-level intervention to enhance the quality of life patient with human immunodeficiency virus

10.11591/ijphs.v13i4.23936
Gusti Ayu Krisma Yuntari , I Made Ady Wirawan , Ni Nyoman Sri Budayanti , Putu Eka Arimbawa
Human immunodeficiency virus (HIV) and acquired immunodeficiency syndrome (AIDS) are still challenges worldwide. Improvement in quality of life (QoL) is crucial for HIV-infected people. The QoL of HIV-infected patients has been improved by the use of multilevel interventions in various circumstances. This study evaluates the evidence for interventions that target the quality of life of persons with HIV. A systematic review was performed according to the PRISMA guidelines. Science Direct, the Cochrane Library database, and Pubmed were also searched. Three independent reviewers extracted the data. Searches were conducted for articles published from 2013 to 2023. Searching procedures and data abstraction techniques were standardized. The 27 research examined the effects of interventions on HIVinfected people's quality of life. China is the most frequently mentioned country in the 27 studies, appearing as the setting in 8 (29.6%). Following South Africa (14.8%), New York (11.11%) and Kenya (7.4%) were the next most often mentioned countries. For analysis purposes, the interventions used in the 27 reviewed studies were classified into three main categories: individual intervention, family-community intervention, and health service intervention. Significant intervention innovations in quality-of-life research combining several interventions, including individual, family, community, and health system intervention, show the most promising results
Volume: 13
Issue: 4
Page: 1757-1768
Publish at: 2024-12-01

Batak Toba language-Indonesian machine translation with transfer learning using no language left behind

10.11591/ijaas.v13.i4.pp834-843
Cevin Samuel , Irsan Taufik Ali
This study focuses on neural machine translation (NMT) for low-resource languages (LRLs) pair, Batak Toba-Indonesian (bbc↔ind). The Batak Toba language is a critically endangered dialect of an Indonesian ethnic group, Batak. Recent advances in machine translation offer potential solutions, with transfer learning emerging as a promising approach for this language pair. We used a publicly available bbc↔ind parallel corpora from the Hugging Face datasets hub and employed the NLLB-200's distilled 600M variant model as the baseline model. Our models achieved sacreBLEU scores as follows: i) for bbc→ind, it achieved a score of 37.10 (+25.67, up from 11.43) and ii) for ind→bbc, it achieved a score of 30.84 (+25.82, up from 5.02). These results outperform all previous works in the task bbc↔ind machine translation and prove the validity of our approach.
Volume: 13
Issue: 4
Page: 834-843
Publish at: 2024-12-01

A classification model for predicting course outcomes using ensemble methods

10.11591/ijece.v14i6.pp7090-7102
Emad Al-Momani , Ala'a Shatnawi , Mohammad Almomani , Ammar Almomani , Mohammad Alauthman
Educational data mining has sparked a lot of attention in latest years. Many machine learning methods have been suggested to discover hidden information from educational data. The extracted knowledge assists institutions in enhancing the effectiveness of teaching tactics and the quality of education. As a result, it improves students' performance and educational outputs overall. In this paper, a classification model was built to classify students' grades in a specific course into different categories (binary and multi-level classification tasks). The dataset contains features related to academic and non-academic information. The models were built using a variety of machine learning algorithms: decision tree (J48), support vector machine (SVM), and k-nearest neighbor (K-NN). Furthermore, ensemble methods (bagging, boosting, random subspace, and random forest) which combined multiple decision tree classifiers were implemented to improve the models' performance. The data set was modified under two stages: features selection method and data augmentation using a method called synthetic minority over sampling technique (SMOTE). Based on the results of the experiments, it is possible to predict the students' performance successfully by using machine learning algorithms and ensemble methods. Random subspace obtained the best accuracy at two-level classification task with modified data with 91.20%. At the three-level classification task, the best accuracy was obtained by random forest with 87.18%.
Volume: 14
Issue: 6
Page: 7090-7102
Publish at: 2024-12-01

Algae content estimation utilizing optical density and image processing method

10.11591/ijece.v14i6.pp6248-6257
Muhammad Wafiq Kamaluddin , Agus Indra Gunawan , Setiawardhana Setiawardhana , Bima Sena Bayu Dewantara , Era Insivitawati , Anja Asmarany , Ariesa Editya Pratama
One of the factors that influence shrimp cultivation is the presence of algae. Precise knowing algae content in the pond is essential for effective management. Most research in the field of algae species carried out by researchers were observing Chlorella Sp. more than the other algae species, with a particular emphasis on substance concentrations. This study proposed non-invasive techniques for quantifying algae abundance, utilizing optical density (OD) and image processing (IP) methods. Three different algae species are frequently found in Indonesia i.e., Chlorella Sp., Thalassiosira Sp., and Skeletonema Sp. are used as sample. Those samples are cultured and prepared in a certain volume with a certain quantity. For experimental and observation purposes, those samples are then diluted into water based on percentage value. The experimental results provided RGB values, which were then used to establish polynomial equations. To verify these equations, two approaches were employed: synthetic image analysis and evaluation using additional data. The mean average error (MAE) was found to be 3.467 for IP method and 3.513 for OD method. It shows that IP method give better result compared to OD method in this study. However, it is very possible that the two methods will complement each other.
Volume: 14
Issue: 6
Page: 6248-6257
Publish at: 2024-12-01

Improving the risk profile of Indonesian enterprise taxpayers using multilabel classification

10.11591/ijai.v13.i4.pp4323-4333
Teguh Prasetyo , Budi Susetyo , Anang Kurnia
Optimizing tax revenues is difficult in Indonesia due to obstacles such as tax evasion and tax avoidance. It is closely related to an organization's compliance with tax regulations, known as the taxpayers risk profile. However, this mechanism does not accurately detect tax avoidance and tax evasion risks. To overcome this limitation, we use a multilabel classification machine learning method in this study, which classifies a single observation into one or more labels at once. The approach involves problem transformation (binary relevance and label powerset), algorithm adaptation (multilabel k-nearest neighbor (ML-kNN) and multilabel-adaptive resonance associative map (ML ARAM)), and ensemble (label space partitioning and random k-label sets with disjoint (RAkELd)). Based on the model performance comparisons, we discovered that the ML-ARAM method based on deep learning is the best, with an average F1-score of 95.5% and a hamming loss of 7.4%. We also examine the feature importance of the best model to reduce the dimensions of features so that we can identify the dominant factors that encourage a taxpayer entity to engage in tax avoidance or tax evasion. The findings of this study improve the accuracy of tax avoidance risk detection and tax evasion risk profiles using machine learning methods, ensuring maximum tax revenues in Indonesia.
Volume: 13
Issue: 4
Page: 4323-4333
Publish at: 2024-12-01

Optical laser-generated electricity for powering tilt-meter sensor

10.11591/ijece.v14i6.pp6140-6147
Nelfyenny Nelfyenny , Dwi Bayuwati , Suryadi Suryadi , Irwan Rawal Husdi , Imam Mulyanto , Aditya Dwi Prasetio , Dedi Irawan , Bambang Widiyatmoko , Andi Setiono
This research investigated the feasibility and efficacy of power over fiber (PoF) transmission systems for geotechnical monitoring applications, addressing challenges associated with traditional power transmission methods. Leveraging fiber optic technology, PoF systems offer advantages such as high reliability, minimal signal loss, and immunity to environmental factors. The study presents a detailed design and implementation of a PoF transmission system, integrating a high-power laser source (HPLS) and photovoltaic technology for efficient power transmission over extended distances. Results demonstrate impressive volt-ampere characteristics and conversion efficiencies, with the optimized system configuration achieving a peak power output of 682 mW. Furthermore, the study evaluated the performance of a surface inclinometer sensor powered by the PoF system, showcasing its effectiveness in monitoring soil movements with remarkable stability and consistent power supply. Future research directions include scalability studies, optimization of system efficiency, and field deployments to broaden the applicability of PoF technology in geotechnical monitoring, ultimately advancing disaster mitigation and infrastructure resilience efforts.
Volume: 14
Issue: 6
Page: 6140-6147
Publish at: 2024-12-01

The study of requirements for the workforce of the digital industries using web scraping techniques

10.11591/ijaas.v13.i4.pp864-881
Rachawit Tipsena , Wirapong Chansanam , Kulthida Tuamsuk , Tossapon Boongoen
This study investigates workforce requirements in Thailand's digital industries, focusing on qualification requirements across five industry groups: hardware and smart devices, software and software services, digital service, digital content, and telecommunication. Employing Python-based web scraping from selected job websites during 2022–2023, the data undergoes natural language processing (NLP) for analysis. Within Group 1 (hardware and smart devices), electrical engineers dominate with 92 positions, emphasizing a demand for engineering expertise. Group 2 (software and software services) sees a surge in programmer roles, totaling 244 positions, showcasing a need for robust programming skills. Group 3 (digital service) prioritizes information technology (IT) support, claiming 354 positions, indicating high demand for IT support qualifications. Graphic design leads Group 4 (digital content) with 587 positions, highlighting the need for a workforce in digital content production. In Group 5 (telecommunication), network engineers dominate with 37 positions, signaling a demand for top-tier network engineering skills. Most positions across groups specify a bachelor's degree and often require prior experience, highlighting the industry's preference for both academic and practical qualifications. This study underscores the digital industry's rapid growth and the sustained demand for a qualified workforce, emphasizing the importance of academic credentials and specialized skills for future employment in these sectors.
Volume: 13
Issue: 4
Page: 864-881
Publish at: 2024-12-01
1 2 3
...
2097 2098
Show 400 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