Journal of Advanced Research Design

Ontology Model for Hajj using Ontology Development 101 Methodology

2025-02-25T09:21:53+00:00

Hajj and Umrah are fundamental spiritual journeys in Islam, requiring proper knowledge and preparation. However, information related to these rituals is scattered across historical texts, religious scriptures, and online sources, making it difficult to access and understand. The challenge of structuring and organizing knowledge is not unique to religious studies but extends to other domains, such as mental health awareness and Islamic thought. With the advancement of technology, particularly artificial intelligence (AI) and ontology-based systems, structured knowledge representation has become increasingly valuable. This study develops a Hajj ontology as a structured knowledge base to systematically organize pilgrimage-related information. The Ontology Development 101 (OD101) methodology was adopted, consisting of specification, integration, conceptualization, implementation, and evaluation phases. The integration phase combined elements from Umrah ontology and Qur’an ontology, ensuring a more comprehensive representation of pilgrimage knowledge. The developed ontology was successfully implemented due to the simplicity and flexibility of the methodology. Protégé was used for visualization, while SPARQL queries enabled knowledge retrieval, addressing competency questions and enhancing accessibility. This structured approach demonstrates the potential of ontology-based models to improve understanding and decision-making in religious studies. Beyond Islamic studies, ontology-based frameworks have broader applications. In mental health research, structuring knowledge on psychological well-being, such as stress, anxiety, and depression, can support more effective awareness strategies. Similarly, in Islamic thought and AI-driven knowledge systems, ontology models help organize and interpret religious teachings, ensuring alignment with Islamic principles. This study highlights how ontology can bridge the gap between fragmented information sources and structured knowledge representation, offering a foundation for further research in religious studies, mental health awareness, and AI-driven knowledge management.

The Influence of Augmented Reality(AR) Technology on Sculpture Art Education

2025-02-05T15:26:49+00:00

Sculpture in the Visual Arts Education subject is one of the topics that is often said to be difficult to learn due to the lack of effective learning tools. Based on this background, this study was conducted to develop and test the effects of using a learning tool based on Augmented Reality technology called the E-Seni Arca Toolkit on the understanding of Form One secondary school students on the topic of sculpture. This study used a quantitative method involving a quasi-experimental study design. The study sample consisted of 50 Form One students from Sekolah Menengah Kebangsaan Mergong, Alor Setar, Kedah. They were randomly divided into a treatment group and a control group. The students in the treatment group used the E-Seni Arca Toolkit to learn the topic of sculpture, while the students in the control group used the conventional method. The study instrument was a set of test questions to measure students' understanding of aspects related to sculpture. An independent t- test was used to analyze the difference in mean scores of the test results between the two groups. The findings showed that the understanding of students in the treatment group about sculpture was higher than the understanding of students in the control group. Therefore, the E-Seni Arca Toolkit can be used as a learning tool in the teaching and learning process to help secondary school students learn sculpture topics more effectively

Do I Need to Consult a Doctor when Making an AMD?

2025-09-25T22:36:03+00:00

Advance Medical Directive (AMD) is a critical tool for ensuring patients’ healthcare preferences are respected when they cannot communicate their decisions. Despite their importance, many individuals are uncertain whether consulting a doctor is necessary when creating an AMD. This study explores the question: “Do I need to consult a doctor when making an AMD?” AMD has gained prominence in modern healthcare because it upholds patient autonomy and reduces ethical dilemmas during medical emergencies. However, the absence of clear guidelines on the necessity of medical consultation poses a challenge. The study aims to assess the impact of involving medical professionals on the quality and comprehensiveness of AMD and identify barriers to seeking such consultations. A qualitative method approach was employed. The understanding of medical terminology, specificity of directives, and perceived confidence in decision-making, while the data from secondary sources explored the perceived benefits and obstacles of involving a doctor. Results reveal that consulting a doctor enhances individuals’ comprehension of medical terminology and treatment options, resulting in more detailed and practical AMD. Participants who sought medical advice expressed greater confidence in their decisions, while those who did not cited concerns such as cost, accessibility, and time as barriers. Healthcare providers emphasized the importance of aligning AMD with realistic medical scenarios, which can be challenging without expert input. The study concludes that although consulting a doctor is not legally mandatory, it significantly improves the effectiveness and reliability of AMD. To address accessibility challenges, integrating low-cost or free medical consultation services into the AMD creation process is recommended. Such measures can ensure broader access to professional guidance, resulting in improved AMD quality and better alignment with individual healthcare goals. This approach can enhance patient autonomy while fostering more effective healthcare outcomes.

Leakage-Aware Research Design for Interpretable PM2.5 Prediction Across Urban Air-Quality Monitoring Sites

2026-06-21T23:11:21+00:00

Air-quality prediction studies often report optimistic model performance when temporal and spatial dependencies are not handled explicitly. This study develops a leakage-aware research design for next-hour PM2.5 prediction using the public Beijing Multi-Site Air Quality dataset. Lagged pollutant, meteorological, calendar, wind-direction, and station features were evaluated with persistence, Ridge regression, random forest, and histogram gradient boosting models. The design compared random, chronological, rolling-origin, station-holdout, multi-horizon, feature-ablation, leakage-stress, and stratified-error evidence with MAE, RMSE, R2, median absolute error, and bootstrap confidence intervals. Using 408,172 feature rows, the best chronological model was HistGradientBoosting with MAE 15.61 ug/m3 and RMSE 29.01 ug/m3, while the best random-split result was 14.90 ug/m3 MAE. The 24-hour horizon increased HistGradientBoosting MAE to 53.03 ug/m3, and a deliberately leaky target feature reduced apparent MAE to 0.85 ug/m3, demonstrating why leakage diagnostics are necessary. The workflow provides a reproducible blueprint for environmental machine-learning studies rather than a new forecasting algorithm.

Enhancing VoLTE Quality using Unsupervised Learning

2025-08-25T20:34:48+00:00

The integration of machine learning (ML) algorithms with data analytics has become essential for optimizing telecommunication services due to the growing complexity of networks. Voice over LTE (VoLTE) is among other technologies that increased this complexity. Hence, optimizing VoLTE to ensure the highest quality-of-service (QoS) attained by end users presents a significant challenge. In this paper, we propose a framework that links key performance indicators (KPIs) with drive test data to gain more insights about the network utility and better manage its resources. The numerical evaluation for our proposed framework demonstrates 25% improvement in the VoLTE QoS compared to other existing approaches.

A Comprehensive Assessment Method Based on the Operation and Resources of a Low-Carbon Cloud Platform

2025-04-02T23:41:21+00:00

In the context of rapid globalization and informatization, cloud platforms have become a critical tool for enterprises to achieve efficient and precise operations, demonstrating significant practical value. This study proposes a comprehensive evaluation method for cloud resources by establishing a full lifecycle data management mechanism. This approach achieves systematic optimization from the infrastructure layer to the business application layer. Specifically, this paper constructs a comprehensive verification system that covers data collection, processing, analysis, and visualization. Additionally, this paper establishes a cross-layer comprehensive evaluation model spanning the platform and application layers, supporting real-time monitoring of operational differences in heterogeneous systems. Through empirical deployment in 27 regions from January to December 2023, we continuously monitored the operational stability and resource efficiency of cloud systems. Experimental results indicate that this method not only enhances the overall management efficiency and resource utilization of cloud platforms but also provides quantifiable technical support for enterprises to gain strategic advantages in digital competition while ensuring system stability

The Integration of Blockchain Technology in Halal Supply Chain Management HSCM): Enhancing Transparency in Islamic Business Practices

2025-09-25T22:37:12+00:00

In the contemporary global halal industry, ensuring transparency, accountability, traceability, and trustworthiness across the supply chain is of paramount importance. However, the industry continues to face challenges related to fraud, mislabeling, and lack of real-time verification. This study explores the integration of blockchain technology into halal supply chain management as a strategic approach to strengthen Islamic business ethics and operational integrity. Blockchain, with its immutable ledger and decentralized system, offers a reliable platform for recording and verifying every transaction and process within the halal supply chain from sourcing raw materials to final consumer delivery. The research adopts a qualitative approach through case studies and expert interviews within halal certification bodies, logistics providers, and halal food manufacturers. It investigates the extent to which blockchain can address key challenges such as fraudulent halal claims, non-compliance risks, and lack of real-time traceability. Findings suggest that blockchain integration not only enhances the efficiency and reliability of halal certification processes but also aligns closely with the principles of amanah (trust), taqwa (good deed), and mas'uliyyah (accountability) in Islamic business conduct. This paper contributes to the growing body of knowledge on Islamic business practices by highlighting how the implementation of blockchain technology can significantly enhance transparency, reduce fraud, and streamline processes within halal supply chains. It offers practical recommendations for halal industry stakeholders, policymakers, and tech developers to leverage this innovation in upholding Shariah-compliant standards while fostering ethical practices and supporting the growth and sustainability of the halal economy.

Reduction of PAPR using Huffrith Algorithm with Different Digital Modulation in NOMA System

2026-06-18T02:15:43+00:00

More efficient radio access approaches are needed for a full-scale deployment of sixth generation (6G) wireless systems because of the huge interconnectedness and rising needs of different services and data-hungry applications. Regarding this, academics and industry have recently been paying more and more attention to non-orthogonal multiple access, or NOMA. Because NOMA has a higher spectrum efficiency than orthogonal multiple access (OMA) approaches, it is suitable for 6G and beyond. Since OMA plan is unable to adequately enable huge connectivity such as IoT, NOMA has been regarded as a vital scheme. In this paper, QAM and APSK digital modulation techniques have been examined, utilizing 256 and 512 orders, respectively. As a result, APSK improves power efficiency rather than negatively impacting it. Huffrith is a new formulation of algorithm approach that can reduce PAPR more effectively than the conventional scheme. Simulation results show that the NOMA system with the 512-APSK modulation scheme achieves a 9.45 dB PAPR reduction and a 3.57% improvement over the OMA system.

Design and Control of Self-Balancing Electric Motorcycle with Simulink Auto-Tune Pid Controller

2025-04-29T14:16:32+00:00

This project describes the design and control of a self-balancing electric motorcycle with a Simulink PID controller. In this system, self-balancing refers to the ability to balance by itself without any assistance from external forces. The self-balancing electric motorcycle should also be able to bounce back after being given external force. This project draws on the theoretical principles of the inverted pendulum. The inverted pendulum system, unlike many other control systems, is inherently unstable. Due to this condition, a controller has to be designed so that the system reaches stability in the unstable state. This project presents motorcycle modeling, designing, and implementing a controller using a PID controller. PID controllers are implemented to minimize error to the least possible value and improve the performance of the controller. The ability to successfully perform self-balancing electric motorcycle is greatly reliant on the proportional gain, (), integral gain (), and derivative gain () values utilized to control the rotation around the front-to-back axis called roll angle and rotation around the vertical axis is called yaw angle. The objective of this research project is to stabilize the motorcycle at speeds less than 10 km/h. The result of the model is compared and analyzed within various , , and values. The response of the system depends on the values of setting time and overshoot percentage. The simulation results show that the controller cannot balance the robot.

The Role of Chemical Engineer in Industry: Review and Computational Bibliometric Analysis Relating to Particle Technology

2025-04-29T11:45:59+00:00

This research analyzed the role of chemical engineers in the industrial sector. Literature review and bibliometric analysis methods using VOSviewer and R Programming were used in this research. This research analyzed the number of studies, country and affiliation distributions, and topic trend analysis. The sample in this study consisted of 73 publications with a publication year range of 2019–2024. The results of the literature review analysis showed that chemical engineers have a crucial role in various industries. Chemical engineers designed and optimized production processes for chemicals, fuels, pharmaceuticals, and a variety of other products. Meanwhile, the results of research trend analysis on articles published in Scopus-indexed journals showed that the number of publications decreased. The decline in publications could be caused by trends in the research world that can change over time. The COVID-19 pandemic also influenced the decline in research regarding chemical engineering in industry. The trending topics in research regarding the themes raised were the American Institute of Chemical Engineers, Accident Prevention, Risk Assessment, Engineers, Chemical Analysis, Chemical Industry, and Students. The present results could help academics determine which problems to research and could be used as a reference for further research.

3D Online Parts Catalog for Quick Identification and Learning

2025-06-09T09:45:33+00:00

People sometimes face challenges when they have to memorize many components or parts of a machine. Although a complete schematic description of a machine and its parts is generally available, many similar parts with different uses can be troublesome, especially for those not specifically in that field. Inventory system managers and item supply custodians must thoroughly understand the parts requested by specific customers, so as not to make mistakes in distributing the ordered parts. Identifying with the correct nomenclature or using part identification number can overcome this problem in general, but it is not without problems. Users who don't understand the code also create the potential for mistaken identification when there are similarities between different items. The general public users experience many difficulties in identifying, selecting, and sourcing suitable parts from the increasingly diverse choices available from various online sources. The ongoing research is developing the concept of an integrated information system cataloging parts using 3D visualization. 3D visualization identifies the parts and shows them in relation to each other in a 3D view derived from a complete machine. The initial prototype was built based on the spiral evolutionary engineering method, as it contained the core features with extended functionality built around them to exchange data, including fetching from external sources. The proposed system is intended to help users speed up learning and identifying each part, thereby reducing mistakes when they need a specific part or to understand it. The catalog presents various data about spare parts, including their availability status, since it can be linked directly to the provider's inventory system or other third parties from which the parts can be purchased. Users can also utilize it to quickly learn about certain parts of the machine since it can be connected to information providers related to those parts, including through the use of AI and automated search engines to gather specific information.

How to Leverage Artificial Intelligence: Based Customer Relationship Management in B2B Marketing using Micro Foundations of Dynamic Capability Approach

2025-03-25T08:43:02+00:00

The transition to Business-to-Business (B2B) models in relationship management and Artificial Intelligence-based Customer Relationship Management (AI-based CRM) is a complex yet definitive process. From the perspectives of both firms and government institutions, AI-based CRM is increasingly regarded as a strategic organizational activity. Both profit-oriented and non-profit organizations are now challenged to implement AI in ways that generate substantial business value. Despite the growing interest, research on the application of AI in B2B marketing—whether for commercial or social purposes—remains limited. To address this gap, this study adopts a dynamic capabilities framework, particularly focusing on micro-foundations within firms and government institutions. Four case studies of Indonesian public and private sector organizations that have integrated AI into their B2B marketing strategies were examined. The key findings highlight several critical intersectoral elements, revealing connections among essential concepts and their impact on overall business value. Additionally, the study suggests that B2B practices in both firms and government offices yield similar outcomes, mediating between technological readiness and the potential of AI-based CRM.

Assessment on Tracking Performance of PID, Cascade Proportional Adaptive Interaction Algorithm and Nonlinear Adaptive Interaction Algorithm Controller of XY Table Ball Screw Driven System

2024-12-30T11:54:03+00:00

Recently, the major focus in machine tools has been on achieving precise positioning, robust monitoring, low-cost manufacturing, and adaptability to disturbances. These recent demands, or paradigm changes, have ushered in a new and complicated era for machining tools and controls. This paper presents a new technique and contribution to compensating cutting force disturbances using the Nonlinear Adaptive Interaction Algorithm (NAIA). The NAIA controller is created by adding a modified nonlinear function to the main Adaptive Interaction Algorithm Controller (AIA) controller. First, the mathematical model of a machine tool for an XY table is identified using a system identification approach based on the system's frequency response function. Secondly, a new control strategy that will provide good tracking performance of the XY table is designed. Then, the proposed technique through simulation using MATLAB/Simulink software and experimental work using a real plant of Googoltech XY table. It successfully demonstrated that the tracking performance of a machine tool was increased significantly through the newly proposed technique that was compared with the basic PID controller. The results indicate that the newly suggested NAIA control approach succeeded in achieving up to 60.2% improvement in comparison with PID (frequency, f = 0.6 Hz) and 53.55% improvement. in comparison with CasPAi (at f = 0.2 Hz). In addition, results showed that the NAIA provides an improvement of 86.29% in terms of Root Mean Square Error (RMSE) for f = 0.6 Hz in comparison with Proportional Integral Derivative (PID) and a 78.68% improvement in comparison with CasPAi.

Designing and Implementing Interactive Digital Module: Innovation for a Technology-Enable Mathematics Literacy in Calculating Heat Transfer

2025-04-29T11:21:34+00:00

Various technology-based products have emerged as an impact of advances in technology reform. Although much research has been conducted to study technology-based products, few studies have developed innovation of interactive module and how these discussions support or hinder a technology-enable mathematics literacy, especially in calculating heat transfer. This research is aimed to design and implement interactive digital module of fractions to fostering a technology-enable mathematics literacy. This research employed design-based research (DBR) to obtain an overview regarding to design and implement the interactive digital module. The subjects of this research were students, teachers, and experts. The results revealed that the interactive digital module was designed and developed using Flipbook design. The contents covered how to do fraction operations completed with materials of fractions, student worksheets, exercise of mathematics literacy tests, interactive videos, and evaluations related to the need of students. The interactive and user-friendly design empower students to explore fractions and relation to the mathematics literacy. According to the validation by experts, the interactive digital module was quite effective in Flipbook design for students. For the implementation, the interactive digital module of fractions can enhance mathematics literacy of students. This research has implications for using interactive digital module as an innovation to improve student’s mathematics literacy of fractions.

Analysing Current Digital Poverty Challenges: Student Perspectives on Technological Device and Internet Access Issues in Kedah Rural Schools

2025-06-14T18:09:39+00:00

In the aftermath of the COVID-19 pandemic, digital poverty has emerged as a critical concern, intensifying the pre-existing disparities in access to information and communication technology (ICT). This article addresses the widening digital divide, particularly affecting economically vulnerable individuals and communities. Post-pandemic, online learning remains a pivotal option, yet economically disadvantaged families may lack the necessary devices and internet connectivity, leading to a stark inequality in learning opportunities. The study focuses on two key aspects: the extent of gadget/device ownership and internet access among informants during online learning. These questions serve as a foundational exploration into the current landscape of digital poverty. Conducted in the context of Malaysia's remote learning (Pembelajaran Dalam Talian or PdPR) enforced during the Movement Control Order (Perintah Kawalan Pergerakan or PKP), the research findings expose significant challenges in facilitating online education. This article endeavours to elucidate these obstacles, offering insights to shape strategies that respond to the evolving demands of technology in a digital Malaysia.

Solution of the Porous Medium Equation with Source Terms via 4-Point Newton-Explicit Group MKSOR Based on Wave Variable Transformation

2025-06-07T10:48:55+00:00

This paper presented the 4-point Newton-Explicit Group Modified Kaudd-SOR (4N-EGMKSOR) iterative method combined with the wave variable transformation to solve the porous medium equation with source terms (PMES). The PMES is a nonlinear heat equation having a variety of naturally occurring physical applications, primarily describing processes involving fluid flow, heat transfer and diffusion. However, finding its exact solution can be difficult. Hence, a numerical method is a good option to obtain the approximate solution for the proposed problem. Firstly, we applied the wave variable transformation to the PMES to reduce the PMES into an ordinary differential equation (ODE). This reduction will significantly minimize the computational complexity of the approximate solution for the PMES. Furthermore, the finite difference scheme was used to discretize the reduced form of the PMES, which leads to a nonlinear finite difference approximation equation. The approximate nonlinear equation was then solved using the Newton method, yielding a system of linear equations. In order to solve the formed system of linear equations, the 4N-EGMKSOR was developed and its formulation was derived. In addition, the effectiveness of the proposed method was examined by performing some numerical calculations and the results were compared to the existing iterative methods, i.e., the Newton-Gauss Seidel (N-GS) and the 4-Point Newton- Explicit Group Modified Kaudd SOR (4N-EGMKSOR). Based on the comparison, the 4N-EGMKSOR iterative method proposed in this work is more efficient in getting the converged solution of the PMES compared to N-GS and 4N-EGKSOR iterative methods.

Prediction of Remaining Service Life of the Yogyakarta-Kaliurang Provincial Road Section of Sleman Regency using the PCI Method

2025-05-31T10:02:26+00:00

Roads, as part of the national transportation system, play a crucial role, especially in supporting economic, social, cultural, and environmental aspects. They are developed through regional development approaches to shape the spatial structure to achieve national development goals. One commonly employed method for assessing road damage is the Pavement Condition Index (PCI). This research aims to predict the remaining service life of the Yogyakarta-Kaliurang Road based on the PCI analysis. This study was based on road damage survey data. The visual observation revealed various types of damage, such as alligator cracking, bleeding, block cracking, bumps and sags, corrugation, depression, edge cracking, lane or shoulder drop-off, longitudinal and transverse cracking, patching and utility cut patching, polished aggregate, potholes, rutting, shoving, slippage cracking, swell, weathering, and raveling. The PCI analysis generated an average value of 59.4%, falling within the fair pavement condition. The highest pavement index was recorded from STA 15+000 to STA 15+100, with a score of 100%, signifying a good condition. In comparison, the lowest pavement index was discovered from STA 19+050 to STA 19+100, with a score of 12%, indicating a serious condition. The analysis of remaining pavement life yielded an average remaining life between 1 to 5 years.

Enhanced Mutual Authentication Scheme for Fog Computing using Blockchain Technology

2025-05-28T10:09:32+00:00

Fog computing is an innovative concept that extends cloud services further and brings their capabilities even closer to end-users by extending them out as far as the network edge. Since edges operate on resources that are closer to the source of data, this goes a process in addressing some of the problems with traditional cloud computing - especially latency. Unfortunately, while this proposal offers some benefits it does not by itself ensure that edge devices are trustworthy or behaving securely. As a result, security continues to be one of the key focus points in fog computing deployment. In this context, authentication is an essential component of any security system. Traditional authentication systems do not work well in the fog computing context; therefore, an efficient mutual i.e. two-way authentication process should be used between edge devices and fog servers because of the low end-to-end latency constraints. This paper proposes an improved mutual authentication scheme specifically designed for fog computing environments to overcome these struggles. Our proposal scheme ensures mutual verification between fog servers and edge devices can effectively strengthen security against potential threats. It makes sure there is very low to no storage overhead on fog servers which increases efficiency, especially in resource-constrained environments. Our scheme is tested through extensive experiments and mathematical analysis.

Assessing Readiness for BIM-AR Adoption in Malaysia Construction Industry: A Study of Causal Relationship

2025-05-07T15:29:12+00:00

This research explores the readiness of BIM-AR adoption in Malaysia construction industry. Three dimensions are proposed: awareness, challenges, and readiness. It studied how these variables affect the construction players on the readiness and willingness to implement BIM-AR in their project. The research hypotheses are: (1) Awareness of BIM-AR influences the Readiness of its adoption and (2) The challenges influence the Readiness of BIM-AR adoption. Participants included 65 construction players which consist of G5-G7 contractors and BIM consultants located in Klang Valley. The questionnaire survey method is applied in this research to analyse the relationship between variables and verify the hypothesis based on the collected 65 valid questionnaires. The partial least square method structural equation model (PLS-SEM) was used to carry out structural equation modelling to study the relationship between latent variables. The model has revealed that increasing awareness and eliminating challenges could help to enhance readiness in adopting BIM-AR. The findings show that, the positive and significant path coefficient between "Awareness" and "Readiness" suggests that a higher level of awareness is associated with increased readiness for BIM-AR adoption. Meanwhile, the negative and significant path coefficient between "Challenges" and "Readiness" indicates that higher perceived challenges are associated with lower readiness for BIM-AR adoption. According to this study, the main challenges impacting the readiness for BIM-AR implementation is the concern among construction stakeholders regarding the associated costs of adopting this technology. The finding also proves that the level of readiness to implement BIM-AR among respondents are high where the highest ranked score of readiness indicates that the respondents are willing to provide training programs and support for BIM-AR implementation. This shows that the majority of respondents are willing to implement BIM-AR in their project. Hence, these findings have provided a valuable insight on the readiness of BIM-AR in Malaysia construction industry.

Performance Ranking of Construction Companies using Data Envelopment Analysis Model

2025-05-14T09:59:20+00:00

Construction sector is an important industry in contribution economy growth of Malaysia. At this presence time, inflation cost is the critical issue that facing by construction sector companies which impacting the financial performance. The purpose of this study is to investigate the performance of 42 construction companies listed in main market of Bursa Malaysia by Data Envelopment Analysis (DEA) model and super-efficiency DEA with duration from year 2013 until 2022. The analysis measurement uses an output-oriented approach which contains the six key financial ratios to identify the most efficient companies. The data that obtained from financial reports through Bursa Malaysia are utilized. The finding of this paper reveals various in efficiency score among the construction sector companies which underscoring the notability of employing the operational efficiency, enhancing construction sector performance for sustainability as well as the ability of competitiveness significantly. This research delivers the valuable findings to the literature on evaluation of efficiency performance within the construction sector, which hold the significant implications for various stakeholders and improvement on their strategic decision making. The future research of this study can be extended to further examine other sector companies such as agricultural product sector, food and beverage sector, and etc. through examining efficiency.

Automated Early Detection of Retinopathy of Prematurity Zones using SWIN Transformer

2025-05-28T10:17:45+00:00

Retinopathy of prematurity (ROP) is known to be the primary cause leading to permanent vision loss in children, which calls for its diagnosis and treatment based on subjective assessment of retinal vascular characteristics; even though this traditional approach is practical, it takes much time and likely results in errors. Therefore, automation is required not only to enhance precision but also productivity. The study proposes an innovative approach to early detection of ROP zones on fundus images between 2015 and 2020. It will use the Swin Transformer model, which has demonstrated superior precision and achieved a performance rate of 90.11%. This work denotes significant advancement in this field, emphasizing the potential of transformer-based architectures for the precise and efficient detection of ROP in clinical environments. The findings underscore the significance of utilizing state-of-the-art, comprehensive learning approaches to enhance early detection procedures, improving clinical outcomes for at-risk newborns.

Comparative study on Ransomware Detection using Machine Learning

2025-07-03T10:11:32+00:00

Ransomware is one of the serious threats to the kind of cyber technology that is emerging, more precisely because of the high level at which businesses are now warming up to platforms on the internet. In most instances, normal traditional approaches to security, among the lot like antivirus and firewalls, fail, given the level of sophistication and high dynamism involved during ransomware attacks. This review is on the application of several machine-learning detection approaches for ransomware using static, dynamic, and hybrid feature sets. The paper compares the accuracy, precision, recall, and F1-scores for different ML algorithms such as SVM, Random Forest, CNN, LSTM, ensemble methods and hybrid models. The result has been remarked that the ensemble method and hybrid models perform better in all aspects than the individual model. These challenges explore not only the diversity of the dataset in ML-based ransomware detection but also the balance between negative/false positives and negatives, including the ability to accommodate new variants of ransomware. Work in this direction would, therefore, indicate that advanced hybrid models, for example, combinations like Transfer Learning with Reinforcement Learning, Autoencoders with Random Forest, and Generative Adversarial Networks with LSTM, are going to contribute toward the improvement of ransomware detection and mitigation to a much greater extent. This study provides valuable insights for future research on ransomware detection, identifying effective techniques and areas for improvement in detection and mitigation.