Real time optimization

Although lead scoring is an essential component of lead management, there is a lack of a comprehensive literature review and a classification framework dedicated to it. Lead scoring is an effective and efficient way of measuring the quality of leads. In addition, as a critical Information Technology tool, a proper lead scoring model acts as an alleviator to weaken the conflicts between sales and marketing functions. Yet, little is known regarding lead scoring models and their impact on sales performance. Lead scoring models are commonly categorized into two classes: traditional and predictive. While the former primarily relies on the experience and knowledge of salespeople and marketers, the latter utilizes data mining models and machine learning algorithms to support the scoring process. This study aims to review and analyze the existing literature on lead scoring models and their impact on sales performance. A systematic literature review was conducted to examine lead scoring models. A total of 44 studies have met the criteria and were included for analysis. Fourteen metrics were identified to measure the impact of lead scoring models on sales performance. With the increased use of data mining and machine learning techniques in the fourth industrial revolution, predictive lead scoring models are expected to replace traditional lead scoring models as they positively impact sales performance. Despite the relative cost of implementing and maintaining predictive lead scoring models, it is still beneficial to supersede traditional lead scoring models, given the higher effectiveness and efficiency of predictive lead scoring models. This study reveals that classification is the most popular data mining model, while decision tree and logistic regression are the most applied algorithms among all the predictive lead scoring models. This study contributes by systematizing and recommending which machine learning method (i.e., supervised and/or unsupervised) shall be used to build predictive lead scoring models based on the integrity of different types of data sources. Additionally, this study offers both theoretical and practical research directions in the lead scoring field.

The use of vapor compression refrigeration (VCR) systems in industrial cooling is essential, but the efficiency of their operation is usually less fortunate because of the rigidity and fixity of the operation that such control systems are utilized in, as it should. The paper addresses how machine learning (ML) and Bayesian optimization (BO) may be combined to improve VCR systems real-time optimization. Optimized floating-point parameterized models of Bayesian Optimized Neural Network (BONN) aims to alter major parameters in the system including refrigerant flow rate and compressor speed depending on the fluctuating environmental and operating conditions. Current study overcomes the limitations of traditional optimization techniques in offering an open-ended, adaptive solution that is found to predict and optimize system performance in different conditions. By so doing, better energy efficiency, cheaper costs of operations, and enhanced reliability of the system are expected to be attained through the research. The experimentation indicates that Bayesian-optimized model with greater accuracy in estimating energy usage and fridge refrigerants staff is substantially better than the traditional models. This paper will help in advancing the cooling technologies which are sustainable because it allows VCR systems to be adaptively controlled in real time.

Battery Electric Vehicles (BEVs) have a high dependency on effective Battery Thermal Management Systems (BTMS) to guarantee better performance, safety and long-life cycle of lithium-ion battery. Such issues associated with temperatures variations demand efficient thermal control systems. The research paper is carried out under the scope of exploring the applicability of machine learning (especially, Support Vector Machines (SVM) and Firefly Optimization Algorithm (FOA)) in optimising BTMS behaviour in real-time. This combination of techniques enables efficiency and flexibility in manipulating cooling systems in the changing conditions of operation, greatly benefiting efficiency and minimizing the requirements of the costly testing facilities. Findings demonstrate that the proposed model manages to predict and optimize cooling power and Coefficient of Performance (COP) with a relatively low cost and scalable solution of thermal management of BEVs.

Ao amigo e orientador Darci Odloak, pela dedicação e incentivo ao longo da realização deste trabalho. À Prof a Míriam Tvrzská de Gouvêa, pela amizade, apoio e pelas sugestões que enriqueceram esta tese. Aos gerentes da área do refino da Petrobras, pela oportunidade de desenvolver novas tecnologias na área da automação industrial. Aos engenheiros Aloísio de Souza Prado e José Geraldo Pinto, gerentes da refinaria Henrique Lage, pelo apoio e oportunidade de implementar o otimizador. Ao Valmir José Camolesi e Luiz Carlos Pivetta, engenheiros de análise de processo da unidade de craqueamento catalítico da refinaria Henrique Lage, pela colaboração na implementação do otimizador. Aos operadores da unidade de craqueamento catalítico, em especial ao Carlos Alberto de Almeida, pelo incentivo e participação nos testes do otimizador. Ao amigo Antenor Cesar Andrade, engenheiro de controle da refinaria Henrique Lage, pelas sugestões, dedicação, incentivo e participação em todas as etapas da implementação do otimizador. Aos amigos Almeida e Lincoln, pela companhia sempre agradável e pelas sugestões e discussões sobre este trabalho. Aos professores, funcionários e amigos do LSCP, pelo excelente convívio pessoal e profissional. Aos meus pais e irmãos, pelo amor, amizade e estímulo na minha carreira profissional. A todos que direta ou indiretamente colaboraram na execução deste trabalho.

Data integration is the process of extracting and combining data from multiple sources to provide a unified and value-added view of information. The problem of data integration is critical for achieving goals such as 360-degree views of products and customers, regulatory compliance, and data-driven decision making. Beyond the data integration engineering components, data quality, data governance, data silos, and other key topics complete the foundations of data integration.

Sensor derived data can add value across the mining operating chain ranging from resource definition, extraction, pre-concentration, mineral process monitoring and assessment of product quality. Most documented studies on sensors in mining focus on specific technologies for specific applications. These studies do not take into account different aims, objectives and operating conditions at different steps in the value chain. The first part of this contribution assesses key physical performance and discriminatory requirements of sensors applied in each portion of the mining value chain. The second part proposes a framework of methods for quantifying the value added by additional sensor information. Integrating the sensor based technology and the economic value quantification allows for both, designing an economically optimal sensor monitoring network along the whole mining value chain and optimizing efficiencies. Illustrative studies demonstrate the significant economic benefits, in p...

To achieve low-carbon, economical, and secure operation of power systems under the "dual carbon" goals, virtual power plants (VPPs) serve as a key technology for aggregating flexible resources such as distributed energy, storage, and demand response, making multi-objective coordinated scheduling critically important. However, the high dimensionality, multiple constraints, and conflicting objectives of this problem pose challenges for traditional optimization methods. Therefore, this paper proposes a hybrid intelligent optimization framework based on an improved Dream Optimization Algorithm (DOA) and the Non-dominated Sorting Genetic Algorithm (NSGA-II), termed DOA-NSGAII, to solve the multi-energy scheduling problem of VPPs with objectives of minimizing economic cost, carbon emissions, and peak-valley load difference. Through customized encoding, decoding, and constraint-handling mechanisms, the framework integrates DOA's global search capability with NSGA-II's multi-objective decision-making advantages. Simulation experiments on the IEEE 14-bus system with three comparative schemes show that, compared with the single-objective economic optimization scheme (S1), the proposed multi-objective coordinated optimization scheme (S3) achieves significant comprehensive benefits-reducing carbon emissions by 30.5% and improving the peak-valley load difference by 12.0%-at the cost of only a marginal 4.9% increase in economic cost. This study validates the effectiveness of the DOA-NSGAII framework in solving complex engineering optimization problems and provides a new approach for multi-objective intelligent scheduling of VPPs.

An application is presented of a new methodology for the real-time operation of ood control reservoir gate control operations of river-reservoir systems to minimize ooding. The methodology is based upon an optimization/simulation modeling approach that interfaces optimization with a one and/or twodimensional unsteady ow simulation model (U.S. Army Corps of Engineers HEC-RAS). The approach also includes a model for short-term rainfall forecasting and the U.S. Army Corps of Engineers HEC-HMS model for rainfall-runoff modeling. Both short term forecasted rainfall in addition to gaged stream ow data and/or NEXRAD (Next-Generation Radar) can be implemented in the modeling approach. The optimization solution methodology is based upon a genetic algorithm implemented through MATLAB. The application is based upon the May 2010 ood event on the Cumberland River system in the USA, during which releases from Old Hickory dam caused major ooding downstream area of Nashville, Tennessee area. One of the major features of the modeling effort and the application presented was to investigate the use of different unsteady ow modeling approaches available in the HEC-RAS) including one-dimensional (1D), two-dimensional (2D), and the combined (1D/2D) approach. One of the major results of the application was to investigate the use the different unsteady ow approaches in the modeling approach. The 2D unsteady ow modeling is based upon the diffusion wave approach was found to be superior for the application to the Cumberland River system

This paper presents an advanced MPC-based energy scheduling framework for islanded microgrids operating under uncertain and dynamic conditions where photovoltaic (PV) generation and energy storage systems (ESS) are integrated, and load management is hierarchically prioritized. The framework employs a hybrid ARIMA and random forest forecasting model to improve day-ahead and intra-day predictions of PV generation and load demand, enabling intelligent demand response, prioritized load shedding, and adaptive storage operation. Moreover, the proposed framework incorporates time-of-use (TOU) pricing and load importance weighting to minimize operational costs while ensuring a reliable power supply for critical loads. Simulation results across four operational scenarios demonstrate that the proposed method achieves approximately 32% improvement in critical load protection, 30% reduction in total operating cost, and 33.3% decrease in total load shedding compared to conventional MPC-based approaches. The proposed approach, therefore, provides a comprehensive, dynamic, and cost-efficient solution for microgrid scheduling and can be extended to multi-microgrid cluster applications in future research.

Control theory deals with disciplines and methods leading to an automatic decision process in order to improve the performance of a control system. The evolution of control engineering is closely related to the evolution of the technology of sensors and actuators and to the theoretical controller design methods and numerical techniques to be applied in real time computing. New control disciplines, new development in the technologies will fertilize quite new control application fields. The status report gives an overview of the current key problems in control theory and design, evaluates the recent major accomplishments and forecasts some new areas. Challenges for future theoretical work are modelling, analysis and design of systems in quite new applications fields. New effective real-time optimal algorithms are needed for 2D and 3D pattern recognition. Design of very large distributed systems has presented a new challenge to control theory including robust control. Control over the ...

No cenário atual, onde a globalização, aliada a um maior nível de exigência por parte do cliente, impõem às empresas um maior empenho por competitividade, a agdidade no desenvolvimento e otimização de produtos torna-se crucial para a sobrevivência das mesmas no mercado. Neste contexto, procurou-se compilar várias técnicas utilizadas em E n g e h dd Q d & em um método integrado para a Ot +o E x p m m i d de MWtwa. Essas técnicas fornecem resultados muito mais rápidos e econômicos do que a tradicional prática de variar um componente de cada vez na mistura, devido ao menor número de ensaios necessários. Entretanto, apesar de não serem tão recentes, as ferramentas aplicáveis à otimização de misturas não têm sido utilizadas pelo seu maior beneficiário (a indústria), provavelmente por falta de divulgação de sua existência, ou, principalmente, devido à complexidade dos cálculos envolvidos. Dessa forma, além do método proposto, desenvolveu-se também um sofiwa~ que irnplementa todas os passos sugeridos, com o intuito de facilitar ainda mais a aplicação dos mesmos por pessoas não especializadas em técnicas estatísticas. Através do software (OptiMix), o método foi testado em uma situação real e em um estudo comparativo com um relato da literatura, a fim de testar sua validade, necessidade de adaptações e consistência dos resultados. A avaliaçio dos estudos de caso demonstrou que o método proposto fornece resultados coerentes com os de outras técnicas alternativas, com a vantagem de o usuário não precisar realizar cálculos, evitando assim, erros e agilizando o processo de otimização.

The integration of artificial intelligence in digital platforms has fundamentally transformed how organizations conduct experimentation and draw causal inferences from user behavior data. This paper presents a comprehensive framework for advanced A/B testing methodologies specifically designed for AI-driven digital platforms operating in the United States market. We examine the evolution from traditional randomized controlled trials to sophisticated causal inference techniques that address the unique challenges posed by machine learning algorithms, personalization engines, and dynamic user interactions. Through empirical analysis of major US digital platforms and case studies from leading technology companies, we demonstrate how advanced statistical methods including propensity score matching, instrumental variables, and difference-indifferences estimators can enhance the reliability of causal claims in AI-mediated environments. Our findings indicate that traditional A/B testing approaches may yield biased estimates when applied to AI-driven systems, necessitating the adoption of more sophisticated causal inference frameworks that account for algorithmic confounding, network effects, and temporal dependencies.

This paper considers a solution of the task of optimal control of a complex processing systems by economical criteria such as cost, profit and output of final product. At the first stage, the optimal values of controlled process parameters are defined by using step by step steady-state optimization method designed by the authors and taking into account features of economical criteria. At the second stage, a maintenance of suggested parameters is implied by using digital minimum variance controller.

Artificial Intelligence is like as human in decision making, understanding the common language and recognizing the speech
of others. This Artificial Intelligence technology is growing rapidly by developing the new algorithms with different tasks. The data
science and machine learning play a major part with in the Artificial intelligence technologies. Computer vision involves the new
development by creating new algorithms where it detects and delivers the perfect object in all fields. In last, 20 years major
improvement where taken place in the Artificial Intelligence technology in many industries. In Artificial Intelligence applications
nowadays majorly, prolog (Programming in Logic) and LISP (List processing) programming language were used. This review paper
shows the overall concepts of Artificial Intelligence which used in all applications and the term data science and machine learning how
it interconnected with the Artificial Intelligence and it shows the improved development with less human working effort in all fields.
This paper shows the real-time face recognition using computer vision using MATLAB.

This paper explores the theoretical application of reinforcement learning (RL) to dynamic resource management in Continuous Integration and Continuous Delivery (CI/CD) environments like build and test environments. Focusing on the scaling and capacity optimization of virtual machine (VM) pools, the study proposes the use of a Deep Deterministic Policy Gradient (DDPG) model, tailored for environments characterized by continuous action spaces and complex, dynamic demands. The paper delineates a theoretical framework where an RL agent dynamically adapts VM allocations based on real-time requirements, potentially enhancing operational efficiency and reducing costs. Tthe research outlines a conceptual model that leverages the capabilities of RL to address resource allocation challenges inherent to modern software development. The discussion anticipates that the integration of RL could revolutionize traditional resource management strategies by providing more agile, efficient, and cost-effective solutions. Future research directions are suggested, focusing on exploration of alternative RL algorithms for practical implementations in CI/CD environments. This work contributes to the literature by proposing a novel approach to optimizing resource management in CI/CD systems, setting a foundation for future studies and technological advancements in the field.

Implementing Zero Trust (ZT) architecture in Critical Infrastructure (CI) is paramount to securing Cyber-Physical Systems (CPS) against sophisticated threats. However, defining granular, least-privilege access policies in these environments is severely hampered by the reliance on heterogeneous, unstructured operational manuals that dictate legitimate procedures. Manual policy derivation from these documents is error-prone, unscalable, and often results in static "over-privileging," violating ZT principles. This paper presents a novel framework utilizing advanced Natural Language Processing (NLP) to automate the extraction of semantic relationships from technical operational manuals to dynamically generate Attribute-Based Access Control (ABAC) policies. We propose a hybrid methodology combining domain-specific Named-Entity Recognition (NER) with Transformer-based Semantic Role Labeling (SRL) to identify actors, actions, assets, and contextual constraints within procedural text. We formulate a mathematical model for mapping extracted semantic triples into formal ZT policy specifications. Simulation results demonstrated a 92.5% F1score in extracting policy-relevant entities and a 78% reduction in time required for access reviews compared to manual baselines. This research provides a scalable pathway for bridging the gap between static documentation and dynamic security enforcement in safety-critical environments.

Implementing Zero Trust (ZT) architecture in Critical Infrastructure (CI) is paramount to securing Cyber-Physical Systems (CPS) against sophisticated threats. However, defining granular, least-privilege access policies in these environments is severely hampered by the reliance on heterogeneous, unstructured operational manuals that dictate legitimate procedures. Manual policy derivation from these documents is error-prone, unscalable, and often results in static "over-privileging," violating ZT principles. This paper presents a novel framework utilizing advanced Natural Language Processing (NLP) to automate the extraction of semantic relationships from technical operational manuals to dynamically generate Attribute-Based Access Control (ABAC) policies. We propose a hybrid methodology combining domain-specific Named-Entity Recognition (NER) with Transformer-based Semantic Role Labeling (SRL) to identify actors, actions, assets, and contextual constraints within procedural text. We formulate a mathematical model for mapping extracted semantic triples into formal ZT policy specifications. Simulation results demonstrated a 92.5% F1score in extracting policy-relevant entities and a 78% reduction in time required for access reviews compared to manual baselines. This research provides a scalable pathway for bridging the gap between static documentation and dynamic security enforcement in safety-critical environments.

In this study, we present a comprehensive framework for integrating MATLAB's image processing capabilities with Programmable Logic Controllers (PLCs) to enhance process control in robotic manufacturing systems. This integration aims to improve automation by enabling real-time defect detection, quality assurance, and adaptive control within industrial processes. We developed a simulation environment in MATLAB to model the interactions between image processing algorithms, PLCs, and robotic systems, allowing for virtual testing and validation. Key results demonstrate the effectiveness of the proposed system in optimizing automation, reducing defects, and increasing overall production efficiency. We also validate the framework through case studies in robotic assembly lines, illustrating the practical challenges and benefits of this approach. This paper concludes with a discussion on future research directions, including the potential for machine learning integration and advanced optimization techniques to further enhance the capabilities of such systems.

This study introduces a comprehensive and cutting-edge predictive model for tuberculosis (TB) incidence, leveraging the
power of Artificial Intelligence (AI) and Machine Learning (ML) techniques, with a focus on Convolutional Neural Networks (CNN).
Implemented through MATLAB, this model aims to significantly improve the accuracy of TB predictions by incorporating diverse and
multi-dimensional data sources and applying state-of-the-art algorithms. The model development involves a thorough process of data
integration, including demographic, environmental, and clinical datasets, to ensure a holistic approach to prediction. The CNN
architecture is meticulously designed and optimized within the MATLAB environment, utilizing advanced layers and activation
functions to enhance model performance. Training protocols include extensive data augmentation and hyperparameter tuning to refine
the predictive capabilities. Validation is performed using rigorous cross-validation methods and a variety of performance metrics such
as accuracy, sensitivity, specificity, and ROC curves, ensuring the model's robustness and reliability. The study also conducts a
comparative analysis of the CNN-based model against traditional statistical models and other ML algorithms, highlighting the
superiority and potential biases of each. The effectiveness of the model is demonstrated through real-world case studies, providing
valuable insights for public health policy and TB control strategies. This transformative approach aims to revolutionize TB prediction
and significantly impact global health outcomes.

Cholera remains a significant public health challenge, particularly in regions with inadequate water and sanitation
infrastructure. Predictive learning and advanced data analytics offer critical tools for anticipating outbreaks and enabling timely
interventions. This article explores the implementation of predictive learning using Convolutional Neural Networks (CNNs) and
MATLAB to predict cholera outbreaks. By leveraging historical data, including cholera incidence rates, meteorological conditions,
and environmental factors, CNNs can recognize complex patterns that signal impending outbreaks. MATLAB provides a robust
environment for data analysis, visualization, and deep learning model development. We detail the steps involved in data collection,
preprocessing, CNN architecture design, training, and evaluation. A case study demonstrates the application of this approach in a high-
risk region, highlighting its potential to improve predictive accuracy, optimize resource allocation, and enhance public health response.
Despite challenges such as data quality and computational demands, the integration of CNNs in cholera prediction presents a
promising direction for mitigating the impact of outbreaks and improving public health outcomes. Cholera prediction, Convolutional
Neural Networks (CNNs), MATLAB, Predictive learning, Epidemiology, Public health interventions.

The incorporation of artificial intelligence (AI) and
deep learning (DL) into IT infrastructure is
revolutionizing corporate operations and
competitiveness in this present-day dynamic digital
economy. These developments are crucial to the
digital transformation of businesses since they
provide relevant capabilities in data processing,
automation, and decision-making. This study
examines the impact of AI and DL on IT
infrastructure, emphasizing how these technologies
can foster business digital transformation. The
findings of this review indicate that by streamlining
workflows, eliminating errors, and automating
repetitive tasks, AI and DL increase productivity and
drive innovation. Through predictive analytics and
modified services, these technologies also enhance
the consumer experience. Furthermore, through
real-time insights and predictive analytics, the
integration of AI and DL into IT infrastructure
improves decision-making capabilities. However,
some associated challenges include concerns about
data security and privacy. In conclusion, the
incorporation of AI and DL into IT infrastructure is
a crucial factor in the digital transformation of
businesses, enabling opportunities for innovation
and growth.

The aim of the study was to give a general account of the ethical principles that need to be infused in the design,
development and deployment (DDD) of AI powered systems. The study was guided by two theories, the deontological theory of ethics
and the descriptive theory. The study used qualitative survey design and convenient sampling was done to get a sample of 20 experts
in various fields related to moral philosophy, AI software engineers, policy makers, end users and legal practitioners. Participants were
conveniently selected to cover a wide spectrum of players involved in the AI development and deployment. The data collection
focused of collecting qualitative data through focused group discussion and expert interviews that were guided by interview schedules.
The findings showed that various stakeholders in the AI industry had concerns related to bias, privacy issues, accountability,
transparency and prejudice. The participants highlighted concerns on privacy issues as important. Participants demonstrated concerns
over the safety of personal data in AI development, data security challenges and propensity for misuse for misuse of personal
information. Participants stressed on need of developing AI systems and testing them to prevent the reprojection of inequalities
through the AI outputs. Other emergent themes were accountability, responsibility for AI generated decisions, lack of clear
frameworks for handling errors or harm. Transparency was repeatedly mentioned as needing explainability in developing trust in AI
products by all stakeholders in ethical their ethical deployment. The research concludes handling of ethical issues in AI development
and use necessitates a multidisciplinary approach that combines technological, regulatory and ethical perspectives. The study
recommends establishing comprehensive rules and regulations to govern AI development, stressing on fairness and inclusivity in the
designing of algorithms alongside formation of functioning mechanisms for oversighting AI industry in achieving accountability and
transparency. This research value is that it advances knowledge of AI ethics further by providing new insights and dimensions from
the diverse perceptions of AI stakeholders. It provides the impetus for including ethical issues in the whole lifecycle of AI systems
which ranges from design, development to deployment. Further investigations need to be done in order to provide a framework for
designing legislation that addresses the intricate ethical issues emergent with AI technologies.

time to help me. Again, I am most grateful to Prof. Duvvuri Subbarao for his support and help in this work. His help and ideas helped me a lot in completing my work on time. I would like also express my gratitude to my best friends and my sincere thanks to postgraduate office. To my beloved family, thanks for their understanding, support and care. They are always close to my heart. Finally, I would like also to thank to all my friends who have been with me throughout my work and to all those who have helped me directly or indirectly, May God bless all of them.

In recent years, the integration of Artificial Intelligence (AI) and Principal Component Analysis (PCA) has significantly transformed drilling engineering, driving notable advancements in both the efficiency and accuracy of subsurface exploration and production. The fusion of these technologies offers a powerful approach to managing and interpreting the vast, complex datasets typically associated with drilling operations. This research looks into the application of AI techniques in conjunction with PCA to analyse well logs, reservoir data, and production metrics, aiming to uncover critical patterns and insights that traditional methods might overlook. By utilizing AI algorithms, particularly machine learning models, this study harnesses the ability of AI to process and learn from large volumes of data, making it possible to predict and optimize drilling outcomes with greater precision. PCA, as a dimensionality reduction technique, plays a crucial role by simplifying these complex datasets, enabling more efficient data processing and enhancing the interpretability of results. The combination of AI and PCA not only streamlines the analysis but also facilitates the identification of key variables and trends that influence drilling performance. Ultimately, this research contributes to the development of more intelligent and data-driven approaches in drilling engineering, promising to optimize operations and reduce risks in subsurface exploration.

The integration of Artificial Intelligence (AI) into education is rapidly transforming how teaching and learning are
conceptualized, delivered, and experienced. Traditional classroom models, which often rely on standardized approaches, struggle to
address the diverse needs of students across varying cultural, linguistic, and socioeconomic backgrounds. AI-powered adaptive
learning platforms offer a compelling solution by leveraging data-driven insights to personalize instruction, optimize pacing, and align
content with individual learner profiles. Through real-time analytics, these platforms can identify knowledge gaps, predict learning
trajectories, and recommend tailored resources, thereby creating more inclusive and responsive learning environments. From a broader
perspective, AI-driven personalization extends beyond academic outcomes to support student engagement, motivation, and retention.
Adaptive platforms incorporate multimodal content ranging from interactive simulations to language-adjusted materials that cater to
different learning styles and abilities. Moreover, continuous feedback loops enable both students and educators to monitor progress
dynamically, fostering greater accountability and informed decision-making. At a systemic level, the use of AI in education has the
potential to reduce disparities by providing equitable access to high-quality resources regardless of geography or institutional capacity.
However, challenges remain in ensuring data privacy, avoiding algorithmic bias, and maintaining the critical role of educators as
facilitators of human-centered learning. By narrowing focus to practical classroom applications, this study underscores the
transformative potential of AI-powered adaptive learning systems in improving student outcomes across diverse classrooms.
Ultimately, AI’s role in education is not to replace teachers but to augment their capacity to deliver personalized, impactful, and
equitable learning experiences.

Traffic offenses pose a significant challenge to road safety, particularly in developing countries like Nigeria, where the consequences often result in severe accidents and fatalities. This paper surveys recent advancements in artificial intelligence (AI) and deep learning methodologies applied to traffic offense prediction. By reviewing various studies, we examine the effectiveness of different models, including multidimensional data analysis, computer vision, and neural networks, in identifying and predicting traffic violations. The literature highlights the importance of integrating diverse data sources and local context to enhance the accuracy of predictive systems. Despite notable progress, significant gaps remain, especially in region-specific applications that consider unique traffic dynamics. Our findings underscore the need for further research to develop robust, context-aware AI solutions that can effectively mitigate traffic offenses and improve overall road safety. This survey aims to provide a comprehensive overview of existing approaches while laying the groundwork for future innovations in traffic management systems.

Analytics-driven IT platforms have become essential for organizations seeking integrated visibility across increasingly
complex and interconnected enterprise functions. At a broad level, such platforms enable the consolidation of data from disparate
operational domains, transforming fragmented information into coherent, real-time insights that support strategic planning and
operational control. As supply chains globalize, financial operations accelerate, and commercial activities become more data-intensive,
end-to-end visibility emerges as a critical determinant of efficiency, resilience, and competitive performance. This article examines the
design of analytics-driven IT platforms that deliver end-to-end visibility across supply chains, finance operations, and commercial
execution. It explores how modern data architectures integrate transactional systems, operational data sources, and external signals to
create unified analytical layers. Within supply chains, analytics platforms support demand forecasting, inventory optimization, and
disruption monitoring. In finance operations, they enable real-time performance tracking, cash flow visibility, and risk assessment. For
commercial execution, analytics enhance pricing strategy, sales performance management, and customer behavior analysis, aligning
market actions with operational and financial realities. The analysis narrows to the architectural and governance considerations
underpinning effective platform design. Topics include data integration strategies, scalability, interoperability with legacy systems, and
the role of advanced analytics and automation in generating actionable insights. Attention is also given to organizational alignment,
data governance, and decision workflows that ensure analytics outputs translate into measurable business outcomes. The findings
highlight that successful analytics-driven platforms are not merely technical solutions but enterprise-wide enablers that connect
strategy with execution. By providing consistent, trusted, and timely insights across core business functions, these platforms strengthen
organizational agility, improve decision quality, and support sustainable value creation in dynamic operating environments.

In Sri Lanka, Leptospirosis is among the leading public health concerns due to its high morbidity and wide spectrum of clinical manifestations. This study uses data from hospitals all over Sri Lanka, from 2016 to 2019, to investigate the main determinants influencing the outcome of Leptospirosis. Statistical methods, such as logistic regression and chi-squared tests, have been used in order to examine the relationship between disease outcome and various clinical and demographic variables. The findings show a remarkable association between certain symptoms, diagnostic variables, and outcomes. Neck stiffness, leukocyte count, and diagnostic methods were the critical predictors established through logistic regression analysis. Although the predictive model has moderate accuracy 43.6% and an AUC of 0.546, it promises individually targeted interventions and better disease management in endemic areas. The present study underscores the importance of evidence-based approaches for the reduction of the leptospirosis burden in Sri Lanka.

AI-based content filtering systems have become essential for moderating the vast and dynamic online space. Widely
adopted by online platforms and governments, these systems promise efficiency in detecting and removing harmful content.
However, their reliance on machine learning introduces critical limitations, including biases in training datasets, lack of
contextual understanding, and challenges in real-time moderation. These issues undermine the effectiveness of content
moderation and raise significant ethical and legal concerns, such as threats to free speech, privacy, and transparency. This paper
explores the limitations and challenges inherent in AI-based content filtering systems and examines the regulatory strategies
needed to address them. The study advocates for a balanced regulatory framework that ensures technological innovation while
safeguarding fundamental human rights by emphasising ethical principles such as fairness, explainability, and accountability.

This paper explores a novel framework for deploying self-optimizing AI agents designed to enforce real-time security policies across dynamic broadband infrastructures. Given the rise of zero-touch networks, increasing traffic heterogeneity, and growing cyber threats, conventional reactive security methods are no longer sufficient. We propose an architecture that combines reinforcement learning (RL), federated observability, and edge-native threat detection. The paper introduces a scalable agent-based model with proactive anomaly detection and self-adjustment capabilities. Key contributions include a hybrid decision loop, a riskweighted policy optimizer, and an adaptive trust index. The proposed solution is validated through simulations and real-world telecom KPIs. The results demonstrate enhanced mean time to detect (MTTD), reduced false positives, and improved threat response efficiency.

Managing projects in complex, multi-stakeholder environments presents significant challenges due to dynamic
requirements, competing interests, and evolving external conditions. Traditional project management frameworks often struggle to
accommodate the fluidity and unpredictability of such environments, necessitating adaptive strategies that balance agility, risk
management, and strategic alignment. This paper explores Adaptive Project Management (APM) as a comprehensive approach that
integrates agile methodologies, risk-based decision-making, and stakeholder engagement to ensure project success. Adaptive strategies
leverage iterative cycles, continuous feedback loops, and flexible resource allocation to respond effectively to changing project
landscapes. Agility plays a critical role in enabling teams to pivot quickly while maintaining alignment with overarching project goals.
However, agility alone is insufficient in multi-stakeholder environments where competing priorities can introduce uncertainty. Risk
management frameworks must be embedded within adaptive strategies to identify, assess, and mitigate both anticipated and emerging
risks in real time. Furthermore, strategic alignment ensures that adaptive practices remain consistent with organizational objectives,
regulatory constraints, and stakeholder expectations. Through a detailed analysis of case studies and best practices, this study provides
actionable insights on integrating adaptive project management techniques into complex environments. Key findings suggest that
successful project execution requires a hybrid approach that blends structured methodologies (e.g., PMBOK, PRINCE2) with flexible,
agile-driven practices. The study also highlights the importance of leadership, collaboration, and digital tools in fostering adaptability.
Ultimately, this research underscores the need for a context-driven project management paradigm, where balancing agility, risk, and
strategy is essential for long-term success in volatile, multi-stakeholder ecosystems.

The integration of artificial intelligence (AI) into healthcare frameworks in the United States is transforming both operational efficiency and administrative effectiveness. As healthcare systems face increasing challenges from rising patient demands, resource constraints, and complex regulatory requirements, AI-driven solutions have emerged as a critical tool for modernization. This paper explores the evolving role of AI across key areas of healthcare operations and administration, offering insights into its transformative potential. From operational perspectives, AI is enhancing clinical workflows, streamlining diagnostic processes, and enabling predictive analytics for patient care. Advanced machine learning algorithms analyse vast datasets to identify patterns, improving disease prediction and treatment personalization. In administrative domains, AI automates repetitive tasks such as billing, coding, and appointment scheduling, reducing errors and improving resource allocation. Tools like natural language processing (NLP) are further optimizing medical documentation and facilitating communication between providers and patients. Despite its potential, the implementation of AI in healthcare faces barriers, including data privacy concerns, ethical considerations, and the need for robust infrastructure. This paper highlights strategies to address these challenges, emphasizing the importance of stakeholder collaboration, regulatory adaptation, and continuous workforce training. The paper concludes with a forward-looking analysis of how AI integration can contribute to a more efficient, patient-centered healthcare system, fostering innovation while addressing the complexities of modern healthcare delivery in the US. By leveraging AI technologies, healthcare frameworks can achieve enhanced operational resilience and deliver improved outcomes for patients and providers alike.

In an era defined by rapid technological evolution, the ability to anticipate future outcomes is pivotal for effective decisionmaking. Predictive insights, derived from advanced data analytics and artificial intelligence (AI), are transforming the way organizations and individuals plan for the future. By analyzing historical trends and real-time data, predictive models enable proactive strategies in sectors ranging from healthcare and finance to manufacturing and urban development. These insights empower stakeholders to identify potential risks, seize emerging opportunities, and optimize resource allocation. The integration of advanced data solutions has redefined the decision-making landscape, moving beyond traditional reactive approaches to a future-oriented paradigm. Techniques such as machine learning, deep learning, and natural language processing (NLP) enhance the accuracy and relevance of predictive models. For example, in the healthcare sector, predictive analytics helps foresee patient outcomes, improving care delivery and reducing costs. In finance, it identifies market trends and fraud patterns, enhancing risk management. This paper narrows its focus to explore how these technologies enable strategic future planning. It examines challenges such as data quality, algorithmic biases, and the ethical considerations inherent in deploying predictive tools. The study highlights how organizations can overcome these obstacles by implementing robust governance frameworks, fostering interdisciplinary collaboration, and prioritizing transparency. Through case studies and real-world applications, this research demonstrates the transformative potential of predictive insights in shaping a data-driven future. It underscores the importance of innovation, adaptability, and ethical practices in harnessing the full power of advanced data solutions for informed and strategic planning.

The integration of Artificial Intelligence (AI) into manufacturing industries is revolutionizing process control and efficiency. AI technologies, including machine learning, predictive analytics, and robotics, offer unprecedented capabilities to analyse vast datasets, predict outcomes, and optimize operations. This transformation addresses the growing demands for higher productivity, costeffectiveness, and sustainability in a competitive global market. By replacing traditional rule-based systems with intelligent algorithms, manufacturing processes can adapt dynamically to changing conditions, reduce variability, and enhance quality. AI-driven solutions significantly improve efficiency through predictive maintenance, real-time monitoring, and process optimization. For example, machine learning models identify patterns in equipment performance, enabling early detection of potential failures and minimizing downtime. Additionally, AI-powered automation enhances operational accuracy, reduces human error, and supports scalability. These advancements are particularly relevant in Industry 4.0, where interconnected systems generate continuous data streams for actionable insights. However, adopting AI in manufacturing is not without challenges. Barriers include high implementation costs, workforce reskilling requirements, and data integration complexities. Overcoming these obstacles demands a structured approach, including investment in AI infrastructure, robust data governance frameworks, and fostering a culture of innovation. This paper explores the transformative role of AI in manufacturing, focusing on its application in process control and efficiency improvement. By analysing case studies and technological advancements, it provides actionable insights for stakeholders aiming to leverage AI for sustainable growth and competitive advantage.

In today’s complex project management landscape, effective risk management and resource allocation are critical to
ensuring project success. Traditional approaches often rely on subjective assessments and static tools that fail to adapt to dynamic
project environments. The advent of data-driven decision-making has revolutionized these processes by integrating advanced analytics,
artificial intelligence (AI), and machine learning (ML) to provide actionable insights and predictive capabilities. This article explores
how data-driven methodologies enhance risk identification, assessment, and mitigation, while simultaneously optimizing resource
allocation to align with project objectives. By leveraging real-time data from multiple sources, such as historical project performance
metrics, economic trends, and stakeholder feedback, project managers can make informed decisions that reduce uncertainties and
maximize resource utilization. Predictive analytics tools enable the anticipation of potential risks and their impacts, allowing for
proactive contingency planning. Furthermore, AI-driven algorithms facilitate resource optimization by analysing project timelines,
budgets, and workforce availability to recommend efficient allocation strategies. Despite its transformative potential, the integration of
data-driven decision-making faces challenges, including data quality issues, lack of standardization, and resistance to adoption. This
article discusses strategies to overcome these barriers, emphasizing the importance of organizational culture, robust data governance
frameworks, and scalable analytics tools. Case studies across various industries illustrate the tangible benefits of adopting data-driven
approaches in managing project risks and resources. By narrowing the focus to practical applications and best practices, this article
highlights the pivotal role of data-driven decision-making in driving project success, fostering agility, and enhancing organizational
resilience in a rapidly evolving business environment.

In today's hyper-competitive and data-intensive business environment, strategic decision-making must evolve beyond intuition and historical analysis. Enterprises increasingly operate in uncertain, complex markets that demand anticipatory and adaptive strategies. Predictive analytics, machine learning (ML), and scenario modeling represent a powerful triad of technologies that, when integrated effectively, transform decision-making from reactive to forward-looking, data-driven processes. This paper examines how organizations can leverage predictive analytics to forecast trends, identify patterns, and quantify risks across key business domains. When combined with ML, these insights are continuously refined through real-time data ingestion and adaptive learning algorithms, improving the accuracy of forecasts and uncovering nuanced relationships that elude traditional methods. Scenario modeling complements these capabilities by enabling decision-makers to test alternative strategies against a range of economic, operational, and market conditions. This multi-layered approach provides a robust foundation for proactive planning, resource optimization, and strategic agility. The integration of these technologies enhances cross-functional alignment, fosters faster decision cycles, and improves resilience in the face of disruption. Use cases from sectors such as finance, logistics, and energy demonstrate how organizations can dynamically evaluate mergers, pricing strategies, supply chain configurations, and risk mitigation plans with greater confidence. The paper concludes with an enterprise-wide adoption roadmap, including data governance principles, workforce upskilling, and cultural readiness for AI-augmented decision environments. By embedding these technologies into the strategic fabric of the organization, businesses can unlock competitive advantage and long-term value creation.

Numerous efforts have been mooted in the attempt to exploit the potential in mHealth towards addressing and tackling the
growing global disease burden. A survey of literature indicates that a significant number of mHealth projects in developing countries
fail the sustainability test, a measure that evaluates the ability of a solution to continuously meet the primary objective. This study
sought to develop a model for evaluating the sustainability of mHealth systems. By adopting the exploratory research design; proposed
parameters for evaluating sustainability of mHealth system are identified through review of various categories of literatures: evaluation
models, mHealth sustainability evaluation models, mHealth experiences in developing countries. Using these factors, a conceptual
model is formulated and statistically validated using Partial Least Squares, a component of Structural Equation Modeling (PLS-SEM)
found in the SmartPLS software. The outcome of the study revealed that sustainability of mHealth systems in developing countries is
influenced by nine factors; Management Factors - Ownership, Technological Factors – System Quality, System Interoperability,
Technology Sustainability, System Relevance, System Scalability and Individual Factors – User Support, System Access and User
satisfaction. The nine factors explained 58.2% of the variance in the sustainability of mHealth system. Ownership, User support,
Technology Sustainability and User Satisfaction had the stronger influence on Sustainability of mHealth. The factors explained
13.25%, 11.09%, 8.7% and 7.01% of the variance respectively. Management factor (ownership) was found to have the greatest
influence on sustainability of mHealth in developing countries. An evaluation score matrix is also proposed.

The U.S. oil and gas sector, a cornerstone of national energy security and economic output, faces mounting regulatory
scrutiny and evolving tax compliance mandates at federal, state, and local levels. Amid increasingly complex reporting structures and
jurisdictional variations, traditional compliance frameworks—often manual, fragmented, and reactive—struggle to ensure tax
accuracy, detect evasive behavior, and optimize revenue recognition. This paper explores the transformative potential of artificial
intelligence (AI) in reimagining real-time tax compliance mechanisms across the upstream, midstream, and downstream segments of
the petroleum value chain. Beginning with a macro-level overview of compliance challenges—ranging from dynamic severance tax
regimes to fluctuating royalty obligations—this study highlights how the industry’s legacy systems are ill-equipped to handle the scale
and velocity of modern regulatory demands. AI technologies, including machine learning, robotic process automation (RPA), and
natural language processing (NLP), offer a paradigm shift by enabling intelligent data ingestion, anomaly detection, audit trail
transparency, and real-time rule enforcement. Through integration with enterprise resource planning (ERP) and volumetric reporting
platforms, AI-driven systems automate filings, flag inconsistencies, and align operational data with tax calculations. The study further
examines use cases and pilot deployments demonstrating AI's impact on curbing underreporting, reducing audit burdens, and
increasing fiscal transparency. The paper concludes with a forward-looking roadmap for public-private collaboration, data
standardization, and regulatory innovation to institutionalize intelligent automation across the sector. In doing so, AI not only enhances
compliance efficiency but also unlocks significant revenue potential for both industry operators and government agencies

The dynamic nature of consumer behavior in digital ecosystems necessitates the adoption of intelligent systems for
marketing optimization. Machine Learning (ML) has emerged as a pivotal technology for transforming digital marketing by enabling
real-time decision-making based on multimodal consumer interaction data. At a macro level, ML empowers marketers to transcend
rule-based targeting by continuously learning from diverse user data streams including click behavior, social media activity, device
usage, sentiment in reviews, and past purchases. This rich interaction data is fused to model customer intent, optimize campaign
delivery, and forecast conversions with high precision. Techniques such as gradient boosting, recurrent neural networks, and deep
learning architectures allow platforms to adaptively segment users and dynamically adjust ad content and timing for maximum impact.
Real-time bidding (RTB) engines also integrate ML models to predict click-through and conversion rates instantaneously, optimizing
ad spend and audience reach. Beyond tactical applications, ML facilitates strategic insights, revealing which channels and messages
drive long-term engagement and return on investment (ROI). Conversion prediction further enhances this framework by evaluating not
only the likelihood of a single transaction but the potential for ongoing loyalty. However, challenges remain in integrating
heterogeneous data sources, ensuring data privacy, and maintaining model transparency. As digital marketing continues to evolve,
ML-based systems offer unprecedented agility and accuracy in campaign execution, reshaping how businesses attract, engage, and
retain customers.

This paper explores a novel framework for deploying self-optimizing AI agents designed to enforce real-time security policies across dynamic broadband infrastructures. Given the rise of zero-touch networks, increasing traffic heterogeneity, and growing cyber threats, conventional reactive security methods are no longer sufficient. We propose an architecture that combines reinforcement learning (RL), federated observability, and edge-native threat detection. The paper introduces a scalable agent-based model with proactive anomaly detection and self-adjustment capabilities. Key contributions include a hybrid decision loop, a riskweighted policy optimizer, and an adaptive trust index. The proposed solution is validated through simulations and real-world telecom KPIs. The results demonstrate enhanced mean time to detect (MTTD), reduced false positives, and improved threat response efficiency.

With the development of artificial intelligence, symptom recognition in the medical field has come into public view; to address issues such as the insufficient robustness of pure dictionary matching and the weak generalization ability of single rule-based systems, this paper designs and implements a multi-module symptom recognition system. By comparing the performance of pure dictionary matching, regular expressions, rule-based methods, and their combined approach, the paper verifies the effectiveness of a hierarchical collaborative strategy in symptom recognition tasks. Specifically, this method enhances the precise matching ability of pure dictionary matching through regular expressions, and improves the matching ability of pure dictionary matching for fuzzy problems through rule-based methods. The experiment adopts a comparative approach, conducting quantitative evaluations from four dimensions-accuracy, precision, recall and processing speed-and concludes with the advantages of the hybrid model.