Asian Journal of Research in Computer Science
Not a member yet
792 research outputs found
Sort by
Classification of Invasive Ductal Carcinoma and Invasive Lobular Carcinoma of Breast Cancer Using the Artificial Neural Network Recurrent Algorithm
No full textAims: The purpose of this study is to classify invasive ductal carcinoma (IDC) and invasive lobular carcinoma (ILC) of breast cancer using the artificial neural networks recurrent algorithm. the use of artificial neural networks Recurrent algorithms can improve the accuracy of breast cancer diagnosis and lead to more effective treatment plans.
Study Design: The method employed is a cross-sectional design.
Place and Duration of Study: The research was conducted in the Computer Laboratory Department of Informatics, Faculty of Mathematics and Natural Sciences, Udayana University, Bali Indonesia.
Methodology: Utilizing physical parameters from mammographic images as input variables for the artificial neural network algorithm.
Results: For Invasive Ductal Carcinoma, the accuracy is 77.5%, sensitivity (recall) is 55%, precision is 100%, F1-Score is 60.97%, specificity is 100%, FPR is 0, and TPR is 0.55. For Invasive Lobular Carcinoma, the accuracy is 77.5%, sensitivity (recall) is 100%, precision is 68.97%, F1-Score is 81.63%, specificity is 55%, FPR is 0.45, and TPR is 1.
Conclusion: The artificial neural network algorithm is capable of classifying Invasive Ductal Carcinoma and Invasive Lobular Carcinoma effectively
Autocheck: Inspection Service Application, Purchase Assistance, and Used Car Buying and Selling
No full textCars are one of the most popular modes of transportation in Indonesia. The high demand for new cars is often accompanied by economic considerations that drive consumers to switch to used cars as a more affordable alternative. However, buying a used car carries risks, especially regarding the condition of the car, which is not always known for certain. The limited knowledge of consumers often forces them to take the car to a workshop for inspection, which is time-consuming and less efficient. To address this issue, the Auto Check application was developed to make it easier for consumers to check the condition of used cars through inspection services. This application provides features for booking inspection services, order status notifications, and transaction history, aiming to offer comfort and efficiency in the used car purchasing process. With easily accessible inspection services, this application not only provides certainty for buyers but also enhances trust between used car sellers and buyers. What sets Auto Check apart from existing solutions is its integration of real-time inspection data and transparency-focused features, which streamline the inspection process and foster trust. The application primarily targets individual buyers and sellers in the used car market, aiming to simplify their transactions and ensure informed decision-making
Assessing the Effectiveness of Cybersecurity Frameworks in Mitigating Cyberattacks in the Banking Sector and its Applicability to Decentralized Finance (DeFi)
No full textThis study evaluates the effectiveness of cybersecurity frameworks in mitigating cyber threats in traditional banking while assessing their applicability to Decentralized Finance (DeFi). Using financial sector reports, cybersecurity incident databases, and DeFi security audits, we analyze compliance with NIST CSF, ISO/IEC 27001, and PCI-DSS alongside factors such as bank size, IT security investments, and regulatory fines to determine their impact on cyber resilience. Logistic regression results indicate that compliance with cybersecurity frameworks reduces cyberattack likelihood (p = 0.0689, marginally significant), while larger institutions face fewer threats (p = 0.0256, statistically significant). However, increased IT security budgets paradoxically correlate with higher attack frequencies (p = 0.0385, statistically significant), suggesting larger attack surfaces may offset security investments. In contrast, DeFi faces disproportionately higher smart contract exploits, flash loan attacks, and oracle manipulation, leading to significantly greater financial losses (F = 216.92, p < 0.001, highly significant) than traditional banking cyber incidents. Regulatory compliance and industry collaboration show promise in reducing attack occurrences, with cyber incidents projected to decline by over 40% by 2029 under stricter enforcement. However, traditional frameworks are insufficient for DeFi’s decentralized structure, necessitating AI-driven threat detection, mandatory smart contract audits, secure oracle mechanisms, and adaptive regulatory frameworks. This study highlights the urgent need for tailored DeFi cybersecurity strategies while reinforcing the effectiveness of compliance-driven models in banking. It provides actionable insights for financial institutions, regulators, and cybersecurity professionals seeking to enhance resilience across centralized and decentralized financial systems
A Comprehensive Review of Shortest Path Algorithms for Network Routing
No full textThe rapid development of digital technology and the increasing interconnection of devices have made computer networks indispensable to modern life. Global data movement, communication, and applications like cloud computing, IoT, e-commerce, and smart cities are all made possible by these networks. Routing algorithms particularly shortest path algorithms are crucial for determining the most effective data transmission routes and are largely responsible for the dependability and efficiency of these networks. Because these algorithms maintain stability and reliability while lowering latency, costs, and energy consumption, they are crucial to network operation.
Shortest path problem solving has long relied on fundamental algorithms with origins in graph theory, such as Bellman-Ford and Dijkstra\u27s. Despite their successes, the growing complexity and dynamic nature of contemporary networks have exposed their shortcomings. Advanced approaches, including heuristic, hybrid, and AI-driven methods, have been developed to get around these challenges. Innovations like ant colony optimization and blockchain-based algorithms have improved computing efficiency, security, and adaptability.
The Internet of Things, VANETs, and SDNs are just a few of the domains that use these algorithms; each has specific requirements, like real-time adaptation and energy efficiency. Reinforcement learning and prediction models driven by machine learning have further increased routing efficiency, while simulation tools such as Mininet and OMNeT++ have been essential for evaluating algorithm performance in practical scenarios. As emerging technologies like blockchain and quantum computing become more widely accepted, shortest path algorithms will continue to advance, ensuring their suitability in the rapidly evolving digital environment. This study, which looks at their development, applications, and possible future directions, emphasizes their importance in creating modern networks
Review on Algorithmic Approaches to Solving Knapsack Problem
No full textThe knapsack problem is a classic optimization challenge where the objective is to maximize the total value of items packed into a knapsack without exceeding its weight capacity It comes in several variants, including the 0–1 Knapsack Problem (0-1KP), the Multidimensional Knapsack Problem (MDKP), and the Quadratic Knapsack Problem (QKP). This Review paper conducts a detailed exploration and analysis of algorithmic strategies developed for solving the knapsack problem (KP). The paper delves into various algorithmic approaches, including advanced dynamic programming, heuristic and metaheuristic algorithms like genetic algorithms and simulated annealing. The goal is to provide a comprehensive comparison and evaluation of these diverse algorithmic approaches, examining their performance, efficiency, and applicability in various real-world scenarios. By highlighting the strengths, weaknesses, and recent developments in knapsack problem-solving algorithms, this review aims to guide future research and help practitioners make informed choices
Exploring the Impact, Challenges, and Satisfaction of Online Learning among Nursing Trainees in Northern Ghana
No full textBackground: The use of technology in many fields, including teaching, learning, and computerized assessment within higher education institutions, has increased due to the rapid development of many fields in our current era, mainly information and communication technology, characterized by high speed.
Aim: This research aims to assess the impact of online learning and examination among Nurses’ and Midwives’ Training College, Tamale students.
Methods: The study used a descriptive institutional cross-sectional survey with 293 respondents selected using a stratified sampling technique. A semi-structured questionnaire was transformed into a Google form and used as the main tool for data collection. Data was analyzed using SPSS version 25, and results were presented in tables and figures.
Results: About 81.0% would accept online examinations, and 75.2% of the respondents indicated that writing exams online allows fulfilling courses. The study showed that 82.4% of the respondents agreed that online learning helps them be more productive, 77.9% were satisfied with the learning experience compared to others, and 67.9% indicated that face-to-face learning impacts their learning more. The study showed that 118(40.7%) indicated that they are inability to ask questions and express themselves, 40(13.8%) indicated that they lack instructor or tutor support, financial factors affect the use of online learning and examination, and 84(29.0%) indicated that lack of knowledge on Information Technology (IT). Also, (37.9%) indicated that the availability of required technology and adequate access to the Internet could facilitate online learning and examination (36.6%), and the provision of data (26.6%) could facilitate online learning and examination.
Conclusion: The study found that respondents accept online learning as a method they will indulge in. However, it was revealed that online learning does not impact their academic performance. Also, challenges like the unavailability of the required technology, incompatibility of some phones and laptops, lack of adequate internet access, heavy workload of online courses, and lack of enough skills to learn online exist
Real-Time Ship Detection and Text Recognition Using YOLO-OCR for Smart Port Applications
No full textThis study presents a new real-time computer vision architecture designed for maritime environments that combines YOLO for object detection and PaddleOCR for text recognition. The YOLO algorithm was tuned to identify ships and text regions using a dataset of over 600 annotated photos. Two output layers with good detection accuracy (mAP 0.90, F1-score 0.89) were obtained by removing the smallest detection scale (P3) in order to speed up inference and lower computational complexity. making it appropriate for marine applications with bandwidth constraints. To enhance OCR ‘robustness in low-quality or variable lighting conditions, detected text regions undergo a lightweight preprocessing pipeline consisting of grayscale conversion, contrast enhancement, and noise reduction. The proposed framework enables automated and continuous ship monitoring, thereby supporting compliance verification, port logistics, and security operations in real seaport environments. Furthermore, the architecture demonstrates scalability toward large scale, real-time maritime surveillance systems
Bridging Educational Inequity in Nepal through Explainable AI and Social Theory Integration
No full textThis research seeks to address persistent socioeconomic disparities in Nepal’s education system by integrating explainable artificial intelligence (XAI) with foundational social theories. While enrollment rates have improved, inequities in access, retention, and learning outcomes remain among communities marginalized by caste, gender, and geography. Existing research and policies often depend on outdated statistical approaches and fail to combine social theory with modern machine learning. To overcome this gap, we adopt a mixed-methods design that blends quantitative modeling with qualitative insights from educators, policymakers, and community stakeholders. Using national datasets (EMIS, NLSS), machine learning models such as Random Forest and XGBoost are applied to predict educational disparities. SHAP (SHapley Additive Explanations) is employed to interpret results and highlight the most influential factors. These patterns are further contextualized using Sen’s Capability Approach and Bourdieu’s Cultural Capital Theory, ensuring that findings reflect both structural conditions and lived experiences. The study delivers several policy-relevant outcomes: a resource allocation framework to support equitable distribution, interactive dashboards for simulating policy scenarios, and earlywarning indicators for student dropouts. Importantly, the qualitative component complements the quantitative models, capturing voices and perspectives often excluded from policy discussions. By linking XAI with equity-focused theories, this work contributes to academic debates on educational data science, while also providing actionable tools for policymakers. Ultimately, it supports evidence-based advocacy that empowers marginalized communities and advances a more inclusive education system in Nepal
Enhancing Quality Assurance Practices in Software Development: Application of Agile Methodology
No full textIn the fast-paced and ever-evolving landscape of the software industry, delivering high-quality software has become paramount. The shift from traditional software development models to agile methodologies has been driven by the need to accommodate complex software requirements and dynamic user expectations efficiently. However, integrating quality assurance (QA) practices within agile frameworks presents significant challenges. This study aimed to assess the effectiveness of various QA practices within agile methodologies. The study employed a quantitative research approach, utilising surveys targeting fifty software developers and QA professionals. The results underscored the critical role of continuous integration and testing in maintaining software quality, while also highlighting the need for enhanced QA visibility and influence, optimisation of documentation, and fostering collaboration between developers and QA professionals. Based on the findings, a comprehensive framework for effectively incorporating QA practices into agile development processes was proposed. To address the challenges identified in integrating Quality Assurance (QA) practices within agile development processes, a framework or strategies are proposed. These include embedding QA early and throughout the agile cycle, enhancing QA visibility and influence, enhancing QA visibility and influence, promoting continuous integration and testing, adopting a holistic approach to automated testing and similar other strategies. By implementing these strategies, organisations can enhance the quality of their software products while retaining the agility and efficiency offered by agile methodologies, thus ensuring that software products meet the highest standards of quality and exceed user expectations
ACBS: A Bounded-Suboptimal Multi-Agent Path Finding Solver for Search-Based Problems
No full textMulti-Agent Path Finding (MAPF) represents a critical computational challenge in robotics and logistics, requiring the coordination of multiple agents to reach their destinations while avoiding collisions. Traditional optimal algorithms, such as Conflict-Based Search (CBS), deliver mathematically perfect solutions but demonstrate poor scalability with increasing agent populations. Bounded-suboptimal variants like Enhanced CBS (ECBS) and Explicit Estimation CBS (EECBS) attempt to balance solution quality with computational efficiency, yet encounter significant difficulties in large-scale scenarios. This paper adopts Agile Conflict-Based Search (ACBS), a novel bounded-suboptimal algorithm incorporating goal decomposition, temporal flexibility, multiple conflict resolution strategies, and agile heuristics to enhance scalability. A thorough empirical investigation on established MAPF benchmarks using agent populations from 100 to 2000 has shown that ACBS can yield up to 5× runtime improvements and higher success rates, while maintaining solution quality within a suboptimality bound of 1.2. We have found ACBS to demonstrate excellent performance in dense environments in particular, which positions it as a promising solution for online applications, including warehouse automation, and coordination of autonomous vehicles while maintaining solution quality within a suboptimality bound of while maintaining solution quality within a suboptimality bound of w = 1.2