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

28,188 Article Results

Integrating random forest and genetic algorithms for improved kidney disease prediction

10.11591/ijai.v14.i4.pp2797-2804
Bommanahalli Venkatagiriyappa Raghavendr , Anandkumar Ramappa Annigeri , Jogipalya Shivananjappa Srikantamurthy , Gururaj Raghavendrarao Sattigeri
This work offers a novel method for predicting chronic kidney disease (CKD) by combining random forest (RF) classification with genetic algorithm (GA) to optimize important parameters. The dataset comprises 1,659 patients with 51 clinical parameters. The suggested method emphasizes the optimization of random state values, test size, and essential hyperparameters, such as the number of trees in the forest, the least number of samples needed at a leaf node, and the smallest number of samples necessary to split an internal node. The optimization process is conducted in two stages: the first stage optimizes the random state and test size, while the second stage focuses on hyperparameters. Through extensive simulations over 50 runs, the study demonstrates that the optimized model achieves an accuracy ranging from 0.9451 to 0.9738. The results indicate a maximum increase in accuracy of 2.09%, showcasing the effectiveness of the GA-RF integrated approach in enhancing model performance. This work provides valuable insights into the impact of parameter optimization on machine learning (ML) models, particularly in medical diagnostics, and offers a robust framework for developing highly accurate predictive models.
Volume: 14
Issue: 4
Page: 2797-2804
Publish at: 2025-08-01

Data-driven support vector regression-genetic algorithm model for predicting the diphtheria distribution

10.11591/ijai.v14.i4.pp2909-2921
Wiwik Anggraeni , Yeyen Sudiarti , Muhammad Ilham Perdana , Edwin Riksakomara , Adri Gabriel Sooai
Indonesia is one of the countries with the largest number of diphtheria sufferers in the world. Diphtheria is a case of re-emerging disease, especially in Indonesia. Diphtheria can be prevented by immunization. Diphtheria immunization has drastically reduced mortality and susceptibility to diphtheria, but it is still a significant childhood health problem. This study predicted the number of diphtheria patients in several regions using support vector regression (SVR) combined with the genetic algorithm (GA) for parameter optimization. The area is grouped into 3 clusters based on the number of cases. The proposed method is proven to overcome overfitting and avoid local optima. Model robustness tests were carried out in several other regions in each cluster. Based on the experiments in three scenarios and 12 areas, the hybrid model shows good forecasting results with an average mean squared error (MSE) of 0.036 and a symmetric mean absolute percentage error (SMAPE) of 41.2% with a standard deviation of 0.075 and 0.442, respectively. Based on experiments in various scenarios, the SVR-GA model shows better performance than others. Compares two- means tests on MSE and SMAPE were given to prove that SVR-GA models have better performance. The results of this forecasting can be used as a basis for policy-making to minimize the spread of diphtheria cases.
Volume: 14
Issue: 4
Page: 2909-2921
Publish at: 2025-08-01

Domain-specific knowledge and context in large language models: challenges, concerns, and solutions

10.11591/ijai.v14.i4.pp2568-2578
Kiran Mayee Adavala , Om Adavala
Large language models (LLMs) are ubiquitous today with major usage in the fields of industry, research, and academia. LLMs involve unsupervised learning with large natural language data, obtained mostly from the internet. There are several challenges that arise because of these data sources. One such challenge is with respect to domain-specific knowledge and context. This paper deals with the major challenges faced by LLMs due to data sources, such as, lack of domain expertise, understanding specialized terminology, contextual understanding, data bias, and the limitations of transfer learning. This paper also discusses some solutions for the mitigation of these challenges such as pre-training LLMs on domain-specific corpora, expert annotations, improving transformer models with enhanced attention mechanisms, memory-augmented models, context-aware loss functions, balanced datasets, and the use of knowledge distillation techniques.
Volume: 14
Issue: 4
Page: 2568-2578
Publish at: 2025-08-01

Enhancing touchless smart locker systems through advanced facial recognition technology: a convolutional neural network model approach

10.11591/ijai.v14.i4.pp3262-3273
Abdul Haris Rangkuti , Evawaty Tanuar , Febriant Yapson , Felix Octavio Sijoatmodjo , Varyl Hasbi Athala
As the world recovers from COVID-19, demand for contactless systems is increasing, promising safety and convenience. Touchless technology, particularly public locker security systems that use facial recognition and hand detection, is advancing rapidly. The system minimizes physical contact, increasing user safety. It uses advanced models such as multi-task cascaded convolutional networks (MTCNN) and RetinaFace, FaceNet512, ArcFace, and visual geometry group (VGG)-Face for face detection and recognition, with a combination of RetinaFace, ArcFace, and L2 norm Euclidean or cosine as the most effective distance metric method, where the accuracy reaches 96 and 90%. 'Yourvault', an application demonstrating this efficient security feature, provides notifications for mask detection, facial authenticity and locker status, offering a solution to the problem of convenience and security of public spaces. Future research could investigate the impact of photo age on facial recognition accuracy, potentially making touchless systems more efficient. In general, the application of this technology is an important step towards a safer and more comfortable world after the pandemic. This model approach can be followed up with more optimal facial recognition.
Volume: 14
Issue: 4
Page: 3262-3273
Publish at: 2025-08-01

Impact of batch size on stability in novel re-identification model

10.11591/ijai.v14.i4.pp2724-2733
Mossaab Idrissi Alami , Abderrahmane Ez-zahout , Fouzia Omary
This research introduces ConvReID-Net, a custom convolutional neural network (CNN) developed for person re-identification (Re-ID) focusing on the batch size dynamics and their effect on training stability. The model architecture consists of three convolutional layers, each followed by batch normalization, dropout, and max-pooling layers for regularization and feature extraction. The final layers include flattened and dense layers, optimizing the extracted features for classification. Evaluated over 50 epochs using early stopping, the network was trained on augmented image data to enhance robustness. The study specifically examines the influence of batch size on model performance, with batch size 64 yielding the best balance between validation accuracy (96.68%) and loss (0.1962). Smaller (batch size 32)and larger (batch size 128) configurations resulted in less stable performance, underscoring the importance of selecting an optimal batch size. These findings demonstrate ConvReID-Net’s potential for real-world Re-ID applications, especially in video surveillance systems. Future work will focus on further hyperparameter tuning and model improvements to enhance training efficiency and stability.
Volume: 14
Issue: 4
Page: 2724-2733
Publish at: 2025-08-01

Non-small cell lung cancer active compounds discovery holding on protein expression using machine learning models

10.11591/ijai.v14.i4.pp2815-2825
Hamza Hanafi , M’hamed Aït Kbir , Badr Dine Rossi Hassani
Computational methods have transformed the field of drug discovery, which significantly helped in the development of new treatments. Nowadays, researchers are exploring a wide ranger of opportunities to identify new compounds using machine learning. We conducted a comparative study between multiple models capable of predicting compounds to target non-small cell lung cancer, we focused on integrating protein expressions to identify potential compounds that exhibit a high efficacy in targeting lung cancer cells. A dataset was constructed based on the trials available in the ChEMBL database. Then, molecular descriptors were calculated to extract structure-activity relationships from the selected compounds and feed into several machine learning models to learn from. We compared the performance of various algorithms. The multilayer perceptron model exhibited the highest F1 score, achieving an outstanding value of 0,861. Moreover, we present a list of 10 drugs predicted as active in lung cancer, all of which are supported by relevant scientific evidence in the medical literature. Our study showcases the potential of combining protein expression analysis and machine learning techniques to identify novel drugs. Our analytical approach contributes to the drug discovery pipeline, and opens new opportunities to explore and identify new targeted therapies.
Volume: 14
Issue: 4
Page: 2815-2825
Publish at: 2025-08-01

Using the ResNet-50 pre-trained model to improve the classification output of a non-image kidney stone dataset

10.11591/ijai.v14.i4.pp3182-3191
Kazeem Oyebode , Anne Ngozi Odoh
Kidney stone detection based on urine samples seems to be a cost-effective way of detecting the formation of stones. Urine features are usually collected from patients to determine if there is a likelihood of kidney stone formation. There are existing machine learning models that can be used to classify if a stone exists in the kidney, such as the support vector machine (SVM) and deep learning (DL) models. We propose a DL network that works with a pre-trained (ResNet-50) model, making non-image urine features work with an image-based pre-trained model (ResNet-50). Six urine features collected from patients are projected onto 172,800 neurons. This output is then reshaped into a 240 by 240 by 3 tensors. The reshaped output serves as the input to the ResNet-50. The output of this is then sent into a binary classifier to determine if a kidney stone exists or not. The proposed model is benchmarked against the SVM, XGBoost, and two variants of DL networks, and it shows improved performance using the AUC-ROC, Accuracy and F1-score metrics. We demonstrate that combining non-image urine features with an image-based pre-trained model improves classification outcomes, highlighting the potential of integrating heterogeneous data sources for enhanced predictive accuracy.
Volume: 14
Issue: 4
Page: 3182-3191
Publish at: 2025-08-01

Optimization of hybrid PV-wind systems with MPPT and fuzzy logic-based control

10.11591/ijeecs.v39.i2.pp747-760
Ayoub Fenniche , Abdelkader Harrouz , Yassine Bellebna , Abdallah Laidi , Ismail Benlaria
The growing demand for sustainable and reliable energy solutions has driven the development of hybrid renewable energy systems (HRES) that combine multiple energy sources. This research explores the integration of solar energy and wind energy systems, utilizing permanent magnet synchronous generators (PMSG) for wind energy conversion. PMSGs are gaining popularity due to their high efficiency and ability to operate effectively in variable-speed wind conditions, making them ideal for hybrid systems. The study focuses on optimizing the energy extraction from both PV and wind systems using maximum power point tracking (MPPT) boost converters. The control for the MPPT boost converters is based on fuzzy logic (FL), a method that offers flexibility and adaptability in managing the non-linear and dynamic characteristics of renewable energy sources. A hybrid system consisting of PV, wind energy, and a battery storage system connected to a DC bus is simulated using MATLAB Simulink. The model demonstrates the effectiveness of integrating PV and wind energy with MPPT-controlled boost converters and fuzzy logic control, ensuring optimal energy utilization, stable system performance, and efficient energy storage. This research underscores the potential of hybrid renewable energy systems, showcasing how advanced control strategies can significantly improve the efficiency and reliability of energy generation and storage solutions.
Volume: 39
Issue: 2
Page: 747-760
Publish at: 2025-08-01

Analytical study of a single slope solar still: experimental evaluation

10.11591/ijeecs.v39.i2.pp850-859
M. Bhanu Prakash Sharma , D. Arumuga Perumal , M. S. Sivagama Sundari , Ilango Karuppasamy
Even though water covers the surface of the Earth in three quarters, many nations face shortages of drinkable water due to rapid global population and industrial growth. Solar power emerges as an efficient solution, particularly in hot climates with water and energy scarcity. This research focuses on a practical solar solution known as a solar still, a basic apparatus designed to convert available salty water into potable water. In this study, a single-slope solar still using acrylic material is experimentally analysed, predicting daily distillate production under varying climatic conditions. Using heat and solar radiation, solar distillation offers a simple, affordable, and small-scale approach to clean water production. The solar still, utilizing acrylic sheets as a basin material, minimizes heat losses and enhances water evaporation rates, making it a promising technology for addressing water scarcity issues. The experimental analysis results revealed a distillate output of 420 ml per 0.49 m² per day.
Volume: 39
Issue: 2
Page: 850-859
Publish at: 2025-08-01

Comparative analysis of Cohen-Coon and Ziegler-Nichols tuning methods for three-phase induction motor with speed sensorless control

10.11591/ijeecs.v39.i2.pp885-895
Christian Vieri Halim , Katherin Indriawati
The use of speed sensors in the speed controller of three-phase induction motors affects the reliability of the induction motors. In addition, the drive engine that is often used in industry is a three-phase induction motor. So, speed sensorless control is needed for induction motors to achieve the best performance. This study uses a discrete disturbance observer (D0) as feedback on the speed sensorless control. The controller used in this method is a discrete PI with the Cohen-Coon (CC) and Ziegler-Nichols (ZN) tuning method. The purpose of this study is to obtain a comparative analysis of the CC and ziegler nichols tuning method using a discrete PI on the speed sensorless control scheme with torque load variation. This study was carried out experimentally using an Alliance AY3A-90L4 induction motor. The results show that the CC tuning method is better under parameter efficiency and robustness against disturbance and ZN is better under parameter reliability.
Volume: 39
Issue: 2
Page: 885-895
Publish at: 2025-08-01

CriteriaChecker: a knowledge graph approach to enhance integrity and ethics in academic publication

10.11591/ijeecs.v39.i2.pp973-986
Garima Sharma , Vikas Tripathi , Vijay Singh
Academic writing is an integral part of scientific communities. This is a formal style of writing used by researchers and scholars to communicate critical analysis and evidence based arguments. This work showcased a graph-based approach for scraping, extracting, representing and evaluating the available academic writing forgery detection criteria and further enhancing the model by proposing a set of new age criteria. The proposed work is based on knowledge graphs and graph analytics capable of selecting subset of 16 criteria from the available superset of a cent of criterias provided by Bealls, Cabells, Shreshtha, and Think.Check.Submit, Scopus, and other relevant authors. The process for detecting the influencial parameters consists of 04 phases: dataset preparation, knowledge graph representation and making inferences through graph analytics and evaluation of results. The experimental results are then compared to the retraction database that consisting of information about retracted articles. The work enables the construction of an experiential knowledge graph that effectively identifies influential criteria, enhancing this list by incorporating new age criteria into current influential set and concluding in result by successfully detecting the academic predatory behavior.
Volume: 39
Issue: 2
Page: 973-986
Publish at: 2025-08-01

Date fruit classification using CNN and stacking model

10.11591/ijeecs.v39.i2.pp1373-1383
Ikram kourtiche , Mostefa M. O. Bendjima , Mohammed El Amin Kourtiche
In North Africa and the Middle East, the date is the most popular fruit, with millions of tons harvested annually. They are a crucial component of the diet due to their exceptional content of essential vitamins and minerals, which confer a high nutritional value. The ability to accurately identify and differentiate between date varieties is therefore of paramount importance in agriculture. It is crucial for improving agricultural practices, ensuring harvest quality, and contributing to the economic development of date-producing regions. In this paper, we propose a hybrid method for classifying date fruit varieties based on two stages. In the first stage, we select the two best-performing pre-trained models from six experimented deep learning models, and we concatenate the feature maps extracted from these two models. In the second stage, we apply different classification methods, including artificial neural networks (ANN), support vector machines (SVM), and logistic regression (LR). The performance achieved by these methods is 97.22%, 98.46%, and 99.07%, respectively. Then, with the stacking model, we combined these methods, and the performance result was increased to 99.38%. This result demonstrates the effectiveness of the hybrid model for identifying date fruit varieties.
Volume: 39
Issue: 2
Page: 1373-1383
Publish at: 2025-08-01

Wirelength estimation for VLSI cell placement using hybrid statistical learning

10.11591/ijeecs.v39.i2.pp840-849
Joyce Ng Ting Ming , Ab Al-Hadi Ab Rahman , Nuzhat Khan , Muhammed Paend Bakht , Shahidatul Sadiah , Mohd Shahrizal Rusli , Muhammad Nadzir Marsono
Optimizing wirelength involves predicting the total length of wires needed to connect different components within a chip during cell placement. It is a fundamental challenge in very-large-scale integration (VLSI) of integrated circuit (IC) design, as it directly impacts the overall performance and manufacturability of chips. Accurate wire-length estimation in the early stages of the design process is critical for guiding subsequent optimization tasks. This paper proposes a novel hybrid linear regression wirelength (hybrid-LRWL) method that combines the strengths of existing methods rectilinear Steiner minimal tree (RSMT) for low-degree nets and a statistical learning-based approach for high-degree nets. Additionally, it compares the performance of three well-established wirelength estimation techniques: half-perimeter wirelength (HPWL), rectilinear minimum spanning tree (RMST), and RSMT. The methods were evaluated using the International Symposium on Physical Design (ISPD) 2011 benchmark suite, considering accuracy and computational efficiency. The experimental results demonstrated that the proposed hybrid method achieves superior accuracy, with a mean error of less than 0.05% in total wirelength, closely approximating RSMT results. The proposed method reduces computational time up to 3.6 times faster than traditional RSM-based methods. The results establish a strong framework for accurate and efficient wirelength estimation in VLSI design for modern, high-performance ICs.
Volume: 39
Issue: 2
Page: 840-849
Publish at: 2025-08-01

Systematic literature review of learning model using augmented reality for generation Z in higher education

10.11591/ijeecs.v39.i2.pp1109-1120
Zulfachmi Zulfachmi , Normala Rahim , Wan Rizhan , Puji Rahayu , Aggry Saputra
Higher education is evolving with innovations aimed at enhancing the quality of learning, and one prominent innovation is the integration of augmented reality (AR) technology into the learning process. AR merges real-world and virtual elements in real-time, creating interactive and immersive educational experiences. This technology supports the display and interaction with virtual objects, enhancing engagement and comprehension among students. However, effective integration of AR in higher education faces challenges such as limited technological infrastructure, the need for skilled lecturers, and the adaptation of teaching methods to suit generation Z's learning preferences. Despite their technological proficiency, many educational institutions struggle to optimally implement innovations like AR. This systematic literature review aims to explore and identify an AR-based learning model suitable for generation Z in higher education. Findings suggest that AR technology can significantly enhance learning by offering engaging visualizations and interactive experiences, aligning well with generation Z's characteristics and learning styles. Effective AR implementation requires suitable platforms, such as mobile, desktop, wearable, and projection platforms, each offering unique benefits. By designing AR learning models that cater to generation Z, educational institutions can improve learning outcomes and experiences.
Volume: 39
Issue: 2
Page: 1109-1120
Publish at: 2025-08-01

Creating inclusive UX: uncovering gender-bugs in higher education website through GenderMag’ing

10.11591/ijeecs.v39.i2.pp996-1004
Maria Isabel Milagroso Santos , Thelma Domingo Palaoag , Anazel Patricio Gamilla
Higher education websites serve as service-providing and information-disseminating platforms which may contain gender-related usability issues that affect how male and female users interact with digital platforms. This study applied the gender inclusiveness magnifier (GenderMag) method to identify and assess these gender-specific usability barriers. Researchers conducted cognitive walkthrough sessions using gendered personas, Abi (female) and Tim (male), uncovering key inclusivity bugs aligned to specific cognitive facets-motivation, information processing style, computer self-efficacy, risk aversion, and learning style. Insights from these walkthroughs guided the creation of a structured usability survey, administered to 200 respondents equally divided between males and females, comprising faculty and upper-year BS information technology students. Statistical analysis revealed significant gender differences specifically in information processing style (p=0.0003), emphasizing distinct preferences for content organization and navigation between genders. The integration of usability factors with GenderMag’s cognitive facets effectively pinpointed areas requiring inclusive design adjustments, guiding future efforts to enhance equitable digital interactions in educational environments.
Volume: 39
Issue: 2
Page: 996-1004
Publish at: 2025-08-01
1 2 3
...
1879 1880
Show 37 of 1880

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