Journal of Science & Technology
Not a member yet
224 research outputs found
Sort by
AI-Augmented Predictive Analytics for Proactive Cloud Infrastructure Management
No full textCloud computing environments involvement in advanced management strategies to ensure optimal performance, cost efficiency, and reliability. Predictive analytics based on AI- augmentation is emerged as a transformative approach to proactive cloud infrastructure management which uses machine learning models and deep learning techniques to predict system failures, optimize resource allocation, and enhance security postures. The aim of this paper is to present a complete analysis of AI-driven predictive models, highlighting anomaly detection, fault prediction, workload forecasting, and self-healing mechanisms
Evaluating the Impact of ChatGPT and Advanced Language Models on Enhancing Low-Code and Robotic Process Automation
No full textThis study investigates the transformative potential of integrating advanced language models, specifically ChatGPT, into the realms of Low-Code Development Platforms (LCDPs) and Robotic Process Automation (RPA). As organizations continue to harness the power of automation to streamline workflows and bolster operational efficiency, the synergy between natural language understanding and automation technologies presents a compelling avenue for further advancements. The research explores how ChatGPT, a state-of-the-art language model, can facilitate a more intuitive and user-friendly interaction between developers and Low-Code platforms, thereby democratizing the application development process. The findings of this study contribute to a deeper understanding of the potential enhancements brought about by advanced language models in the context of Low-Code and RPA
Analyzing Time Complexity in Machine Learning Algorithms for Big Data: A Study on the Performance of Decision Trees, Neural Networks, and SVMs
No full textThis research paper presents an in-depth analysis of the time complexity associated with three prominent machine learning algorithms—decision trees, neural networks, and support vector machines (SVMs)—in the context of big data. With the growing influx of large-scale data in various sectors, the ability of machine learning algorithms to process and analyze this data efficiently has become paramount. In this study, we focus on evaluating the computational performance of these algorithms, with particular emphasis on how they scale when applied to big data environments. The paper begins by discussing the theoretical foundations of time complexity and its significance in machine learning, especially in scenarios involving extensive datasets. We highlight the importance of understanding time complexity not only from an algorithmic perspective but also in terms of real-world application where both accuracy and computational efficiency are critical for large-scale deployments.
The decision tree algorithm, known for its simplicity and interpretability, is widely used in various data mining and machine learning tasks. However, when dealing with large datasets, its performance can suffer due to its recursive nature and the need to search through many possible splits at each node. We analyze the time complexity of different types of decision trees, including classification and regression trees (CART) and random forests, to determine their scalability limits. The study examines how decision trees perform under various data distribution patterns and feature dimensionalities, providing insights into how their time complexity grows with increasing dataset size and feature space.
Neural networks, specifically deep learning models, have gained popularity for their ability to model complex patterns in large datasets. Despite their high accuracy, especially in tasks involving unstructured data such as images and text, their time complexity poses significant challenges. This paper provides a detailed analysis of the time complexity of feedforward neural networks, convolutional neural networks (CNNs), and recurrent neural networks (RNNs). Special attention is given to the number of layers, nodes per layer, and the impact of training algorithms, such as stochastic gradient descent (SGD) and backpropagation, on the overall time complexity. The analysis also explores how the increasing size of training data and the depth of neural networks affect computation time and memory usage, ultimately impacting their viability for big data applications.
Support vector machines (SVMs), another widely used algorithm, are known for their strong theoretical foundations and ability to provide high-accuracy results, particularly in classification tasks. However, SVMs tend to struggle with scalability when applied to large datasets, primarily due to their quadratic time complexity in the training phase. This research investigates the computational limitations of SVMs, focusing on both the primal and dual formulations of the algorithm. We analyze the impact of kernel functions, such as linear, polynomial, and radial basis functions (RBF), on time complexity and performance, especially when dealing with high-dimensional data. The study further explores optimization techniques, such as the use of support vector approximation and parallelization, to improve the scalability of SVMs in big data environments.
In addition to the theoretical analysis, this paper provides empirical results based on the implementation of these algorithms on large datasets from various domains, including healthcare, finance, and e-commerce. We compare the computational efficiency of decision trees, neural networks, and SVMs under different big data scenarios, evaluating factors such as dataset size, feature dimensionality, and class distribution. The results of these experiments offer valuable insights into the practical trade-offs between time complexity and model accuracy, enabling practitioners to make informed decisions when selecting machine learning algorithms for large-scale data analysis.
Furthermore, the paper discusses the role of hardware accelerators, such as graphics processing units (GPUs) and tensor processing units (TPUs), in mitigating the computational bottlenecks associated with these algorithms. We explore how parallelization and distributed computing frameworks, such as Apache Spark and Hadoop, can be leveraged to improve the performance of machine learning models in big data contexts. The integration of these technologies with machine learning algorithms can significantly reduce training and inference times, making it feasible to apply computationally intensive models, such as deep neural networks, to massive datasets without sacrificing performance.
The findings of this study contribute to a deeper understanding of the computational complexities associated with decision trees, neural networks, and SVMs, particularly in the context of big data applications. By providing both theoretical and empirical insights, the research offers a comprehensive evaluation of the trade-offs between algorithmic accuracy, computational efficiency, and scalability. Ultimately, the paper underscores the importance of selecting appropriate machine learning models based on their time complexity, especially when dealing with the growing demands of big data. The analysis presented here is intended to guide data scientists, machine learning engineers, and researchers in the development of more efficient and scalable machine learning solutions for large-scale data processing
Cloud Adoption in Large Enterprises: A Case Study on Implementing Enterprise Architecture Frameworks for Seamless Integration
No full textThe integration of cloud technologies into existing IT ecosystems has become a critical objective for large enterprises seeking enhanced scalability, operational efficiency, and competitive advantage. However, the adoption of cloud solutions presents significant challenges, particularly in aligning new technologies with legacy systems and organizational objectives. This case study investigates how large enterprises strategically employ enterprise architecture (EA) frameworks to facilitate seamless cloud integration, emphasizing the role of structured frameworks in managing complexity, ensuring compatibility, and maintaining governance. Enterprise architecture frameworks provide organizations with standardized methodologies and principles, such as The Open Group Architecture Framework (TOGAF), Zachman Framework, and Federal Enterprise Architecture Framework (FEAF), which serve as guides to design, implement, and manage cloud environments in a way that aligns with organizational objectives and existing infrastructures. The use of these frameworks is increasingly critical in large enterprises where cloud adoption impacts various functional areas, ranging from data security to operational agility and cost management.
This study explores the technical and strategic aspects of enterprise architecture frameworks in the context of cloud adoption, highlighting the value of structured planning and governance. The case study examines a large enterprise’s journey through each stage of the cloud integration process, from initial assessment and planning to deployment, optimization, and ongoing governance. Key technical challenges, such as data migration, application interoperability, and compliance with industry standards, are analyzed to provide insights into how EA frameworks can address these issues. The alignment of cloud strategies with enterprise architecture also brings to light critical concerns related to risk management, particularly concerning data privacy, cybersecurity, and regulatory compliance. For example, the study demonstrates how TOGAF’s ADM (Architecture Development Method) enables structured phases of cloud integration, facilitating the identification of core IT assets that require restructuring and highlighting potential integration points for cloud-based services. This approach ensures that the cloud adoption process adheres to established architecture principles, such as modularity, scalability, and reusability, while supporting the organization\u27s broader digital transformation goals.
Further, the case study delves into the practical aspects of employing EA frameworks to bridge the gap between on-premises systems and cloud environments. One of the primary technical considerations is the orchestration of services across hybrid or multi-cloud environments. Here, EA frameworks guide the creation of interoperability standards, allowing the seamless flow of data and functionality across diverse platforms. Additionally, the study addresses how EA frameworks support decision-making processes related to workload distribution, resource allocation, and cost management. For instance, enterprises often leverage EA frameworks to evaluate which applications are best suited for the cloud, ensuring that mission-critical applications maintain high availability and performance standards post-migration.
The case study also addresses the organizational aspects of cloud adoption, specifically how EA frameworks facilitate cross-departmental collaboration and stakeholder alignment. Integrating cloud technology into an existing IT ecosystem is not only a technical endeavor but also one that requires strategic alignment across business units. EA frameworks serve as a unifying language, enabling various departments to communicate and align their objectives, requirements, and resources. By providing a clear, structured approach to cloud adoption, these frameworks enhance the enterprise’s ability to adapt to the evolving technological landscape while minimizing disruptions to daily operations.
Moreover, this study highlights best practices in governance and compliance, underscoring the importance of EA frameworks in establishing robust governance mechanisms that ensure cloud environments operate within defined parameters. Compliance with industry regulations, such as the General Data Protection Regulation (GDPR) and industry-specific standards, remains a priority for enterprises, particularly those in heavily regulated sectors like finance and healthcare. EA frameworks provide the structure needed to define and enforce compliance requirements, leveraging tools and processes that monitor and audit cloud activities across departments. This facilitates real-time insights into compliance status, enabling enterprises to respond swiftly to potential risks
Data Versioning and Its Impact on Machine Learning Models
No full textData versioning in machine learning is of paramount importance as it ensures the reproducibility, transparency, and reliability of ML models. In the dynamic landscape of ML research, where models heavily rely on diverse datasets, data versioning plays a crucial role in maintaining consistency throughout the ML pipeline. By tracking changes in datasets over time and aligning machine learning models with specific versions of data, researchers can reproduce experiments, verify results, and address challenges related to data quality, collaboration, and model training. Effective data versioning practices contribute to the robustness of ML workflows, fostering trust in model outcomes and supporting advancements in the field
Unveiling the Essence of Software Quality Attributes: A Comprehensive Review
No full textThis review article delves into the multifaceted realm of software quality attributes, elucidating their pivotal role in shaping the excellence of software products. Software quality attributes, also known as non-functional requirements, constitute the foundation of a robust and user-centric development process. This comprehensive review explores the significance and impact of key quality attributes, ranging from reliability and performance efficiency to security, maintainability, usability, and scalability. Quality requirements, captured as nonfunctional requirements in the early steps of software development, greatly influence the software system’s architecture [1].
The article scrutinizes how reliability ensures consistent performance, the ways in which performance efficiency optimizes resource utilization, and the critical role of security in safeguarding against cyber threats. It also delves into the importance of maintainability for seamless adaptability, usability for enhancing user experience, and scalability for accommodating growing workloads. Software quality is a critical factor in ensuring the success of software projects. Numerous software quality models have been proposed and developed to assess and improve the quality of software products [2].
By providing insights into the evaluation methods, tools, and best practices associated with each quality attribute, this review equips developers and stakeholders with a holistic understanding of the intricacies involved in crafting high-quality software. The synthesis of theoretical concepts and practical considerations offers a valuable resource for navigating the evolving landscape of software development, emphasizing the need for a balanced approach to achieve software excellence. In the face of global competition, businesses across various industries have increasingly turned to lean methodologies to enhance their production processes and remain competitive [3]. It investigates user personas, mental models, and usability studies to enhance the alignment of system requirements with user expectations and needs [4]
Big Data Analytics-Driven Project Management Strategies: Utilizing Artificial Intelligence for Dynamic Scheduling, Risk Prediction, and Automated Task Prioritization in Complex Projects
No full textThe integration of Artificial Intelligence (AI) and Big Data Analytics (BDA) in project management has become a critical enabler of efficiency in managing large-scale, complex projects. This research paper delves into how AI-driven big data analytics can revolutionize traditional project management methodologies by introducing dynamic scheduling, real-time risk prediction, and automated task prioritization strategies. These advanced techniques, which leverage machine learning (ML) models and extensive historical project data, enable a shift from reactive to proactive project management, ensuring that risks and resource constraints are identified and addressed before they impact project delivery. By analyzing massive datasets, including historical performance metrics, resource availability, and project timelines, AI-driven systems can forecast delays, assess risk levels dynamically, and adapt schedules in real-time. This proactive approach facilitates better decision-making, optimized resource allocation, and improved project outcomes.
The study is anchored on the premise that the sheer volume of data generated in large-scale projects often overwhelms traditional project management systems. By incorporating AI and BDA, project managers can better utilize this data, turning it into actionable insights that inform intelligent decision-making. Machine learning algorithms, particularly those specializing in predictive analytics, are capable of identifying patterns that elude human analysis, allowing for the accurate forecasting of project risks, schedule slippage, and task dependencies. This ability to predict potential issues, such as resource bottlenecks or unforeseen delays, enables project teams to implement mitigative actions in advance, thus reducing the likelihood of project failure.
Furthermore, dynamic scheduling is a key focus of this research, as AI-powered models can continuously adjust project timelines based on real-time data. These models consider variables such as resource utilization rates, task dependencies, and evolving project constraints, offering adaptive scheduling mechanisms that evolve throughout the project lifecycle. The automated task prioritization system, powered by BDA, ensures that the most critical tasks receive the appropriate level of attention at the right time, improving project performance and enhancing resource efficiency. Through natural language processing (NLP) and advanced data mining techniques, AI models can also analyze project documentation and communication channels to detect potential risks and suggest task adjustments.
The paper also discusses the application of AI in risk prediction, focusing on how AI models can analyze risk factors from historical data, including resource constraints, financial limitations, and market volatility, to produce risk profiles that project managers can use for strategic planning. Real-time risk assessments, made possible by the integration of AI and BDA, can help project teams stay ahead of potential disruptions. This allows for more accurate contingency planning and reduces the overall risk to project timelines and budgets.
Practical applications of these AI-driven strategies are presented through case studies of large-scale projects in various industries, including construction, information technology, and healthcare. These case studies demonstrate how AI-powered analytics have been successfully implemented to enhance project efficiency, optimize resource allocation, and minimize risks in complex projects. The study underscores the importance of integrating these technologies into modern project management frameworks to cope with the increasing complexity of projects in today’s fast-paced business environment.
While the potential benefits of AI and BDA in project management are substantial, this paper also addresses the challenges associated with their implementation. One significant challenge is the quality and availability of data required to train AI models effectively. Incomplete or inaccurate data can lead to unreliable forecasts, compromising the project’s success. Additionally, the paper explores the issues of data privacy and security in AI-driven project management systems, highlighting the need for robust data governance frameworks to ensure the ethical use of AI technologies.
Another key consideration is the resistance to change within organizations, where traditional project management methods are deeply ingrained. The paper emphasizes the need for a cultural shift towards data-driven decision-making and suggests strategies for fostering an environment conducive to AI adoption. This includes training project management teams to work alongside AI systems and fostering collaboration between AI experts and project managers to ensure smooth implementation and operation.
Finally, this research outlines future trends in AI and BDA for project management, suggesting that further advancements in AI technologies, such as reinforcement learning and more sophisticated natural language processing algorithms, will drive the next generation of intelligent project management systems. These future systems are expected to be even more adept at handling the complexities of large-scale projects, offering real-time solutions to unforeseen challenges and adapting dynamically to changing project requirements
Agile and DevOps: Elevating Software Quality through Collaborative Practices
No full textThis review article explores the symbiotic relationship between Agile methodologies and DevOps practices and their profound impact on elevating software quality. The amalgamation of Agile\u27s iterative, customer-centric approach and DevOps\u27 automated, collaborative ethos reshapes traditional development workflows, fostering a culture of continuous improvement and rapid, reliable software delivery. Software quality is a critical factor in ensuring the success of software projects. Numerous software quality models have been proposed and developed to assess and improve the quality of software products.
Agile methodologies, known for their emphasis on iterative development, customer collaboration, and adaptability to change, contribute to software quality through early issue identification and alignment with user expectations. On the other hand, DevOps practices, including continuous integration, collaborative cultures, infrastructure as code (IaC), and automated testing, emphasize efficiency, reliability, and end-to-end visibility in the development process. Software development is an organized thrives to deliver products in faster, better and cheaper ways.
The synergy between Agile and DevOps accelerates feedback loops, ensuring prompt issue resolution and continuous improvement. The collaborative culture promoted by both approaches enhances shared responsibility for software quality, breaking down silos between development and operations teams. In the face of global competition, businesses across various industries have increasingly turned to lean methodologies to enhance their production processes and remain competitive.
Continuous integration and deployment pipelines, integral to DevOps practices, facilitate the rapid and reliable delivery of thoroughly tested code. Infrastructure as code ensures consistency and repeatability, minimizing configuration-related issues. Automated testing and monitoring embedded in DevOps practices contribute to proactive quality assurance, detecting issues early in the development process. It investigates user personas, mental models, and usability studies to enhance the alignment of system requirements with user expectations and needs.
The impact of Agile and DevOps extends beyond individual practices, creating a holistic environment that fosters adaptability, collaboration, and automation. This review article navigates through the key principles of Agile methodologies and DevOps practices, highlighting their individual contributions and the collective impact on software quality. Iterative development breaks the project into iterations of variable length, each producing a complete deliverable and building on the code and documentation produced before it.
As organizations increasingly adopt Agile and DevOps, they position themselves to not only meet but exceed user expectations. The intertwining of these methodologies reflects a strategic commitment to innovation, resilience, and excellence in the realm of software development. This exploration serves as a comprehensive guide for professionals navigating the complexities of contemporary software development, emphasizing the importance of collaborative practices in achieving and sustaining elevated software quality. Organizations that navigate this dynamic horizon successfully will be those that embrace change, foster a culture of continuous learning, and leverage technology not just for efficiency but as a catalyst for excellence
Tele-dentistry and Data Science: Enhancing Access and Quality of Dental Care
No full textTele-dentistry, coupled with data science techniques, presents a transformative approach to revolutionize dental care delivery, particularly in enhancing accessibility and quality. This paper investigates the synergistic potential of tele-dentistry and data science in enabling remote consultations, diagnostics, treatment planning, and patient monitoring. By leveraging digital technologies and advanced analytics, tele-dentistry extends dental services beyond traditional clinic settings, overcoming geographical barriers and improving healthcare outcomes. Key areas explored include teledentistry platforms, data-driven decision-making, artificial intelligence in diagnostics, patient engagement strategies, and regulatory considerations. Through a comprehensive analysis, this paper underscores the pivotal role of tele-dentistry and data science in addressing disparities in dental care access and advancing the standard of oral healthcare provision
Life Cycle Assessment of Urban Passenger Transportation
No full textThis paper describes a life-cycle analysis (LCA) performed for urban passenger transport (cars and urban buses) in United States for the years 2015–2050 in five-year steps. It is assumed that the only change made is the propulsion technology/fuel used and that the distance driven by all vehicles (within each mode) per year remains the same. The complete conversion to a single powertrain type is neither realistic or necessarily desirable, but the intention of this LCA is not to provide a forecast of greenhouse gas (GHG) and air pollution emissions or other factors; the system is too complex with more uncertainties. Instead, the intention is to compare the effect of applying the various technologies/fuels in order to discern their effect in a simplified manner and identify any possible limitations and allow further discussions on the ideal policy strategy