Fusion: Practice and Applications

The detection of diseases in hydroponically cultivated saffron should be carried out as early and accurately as possible to maintain the quality of the yield, minimize losses, and promote sustainable farming practices. Manual diagnosis strategies are not suitable for high-density hydroponic systems where early symptoms tend to be subtle, as these methods are slow and rely extensively on experts. This research aims to develop a novel framework based on deep learning technology, using a Diffused Concurrent Convolutional Neural Network (DCCNN) to perform image analysis and detect diseases in saffron crops. The modified DCCNN includes a hierarchical three-stage classification pipeline consisting of crop recognition, disease detection, and Classification of the specific diseases, adding an “unknown” category for non-target or ambiguous outputs at each stage to enhance flexibility. The digression from standard deep learning techniques is justified due to the DCCNN construction, which contains a learnable diffusion layer and concurrent multi-scale convolutional blocks, and thus encapsulates strong feature propagation with fine-grained detection of complex and low-data environments. Evaluation on a specially annotated dataset of hydroponic saffron showed strong performance with up to 99.4% classification accuracy, exceeding well-known CNN baselines including EfficientNet and ResNet50. Additionally, the model processes static crop images and associated environmental sensor data, collectively referred to as 'non-sequential crop data,' focusing on spatial features without temporal dependencies. These findings confirm that the system, which is based on DCCNN, provides a transferable solution for precision disease detection in controlled-environment agriculture systems and can be extended to other high-value crops.

This study aims to evaluate the reality of administrative empowerment in Egyptian universities, with a case study conducted at Al-Arish University, considering national efforts to improve institutional performance and the quality of higher education. Using a descriptive-analytical design, data was gathered through a structured questionnaire distributed to a purposeful sample of 40 administrative staff and mid-level managers. Data gathered from the structured questionnaire were analyzed to determine the levels of administrative empowerment at each of its key dimensions by using descriptive statistical methods comprising frequency distributions and weighted means. The study investigates the effectiveness of empowerment practices and their influence on institutional outcomes. Findings reveal a broad disparity between theory and practice with the levels of empowerment ranging from low to moderate. The major hindrances to participation are poor employee participation in decision-making, a lack of managerial support, discriminatory workplace culture, and low leadership development. Restricted senior management-employee interactions were also reported to hinder participatory practices. The study recommends that empowerment policies with clarity, actionable implementation strategies, and organizational resolve are necessary for creating a positive and motivational work culture. Recommendations made include enhancing internal communication, innovation, and incentive mechanisms to improve administrative performance. Such steps are vital to ensuring a productive environment to facilitate sustainable development and institutional change. Empirical findings demonstrate the significance of strategic empowerment to aid governance reform and quality assurance in Egyptian higher education.

Cloud computing technology offers key advantages for smart grid applications, especially for electric companies with industry expertise. However, implementing cloud-based solutions for smart grids requires careful consideration of several critical factors. This research aims to identify the primary factors influencing electricity companies' the adoption of cloud-based solutions in smart grids in Oman. An in-depth interview was conducted with field experts to develop a comprehensive model that will potentially be a key reference source for guiding the adoption and implementation of cloud-based smart grids in Oman and beyond. This research is espoused by the technology organization environment (TOE) framework and the diffusion of innovation (DOI) theory. The model identifies ten key factors that impact the adoption of smart grid cloud-based (SGCB) solutions in utilities in Oman. By understanding the significance of these factors, utility companies can make well-informed decisions about implementing cloud-based solutions for smart grids. This research serves as a valuable resource, guiding the adoption of this technology in the electricity sector. It also contributes to smart grid advancement and optimization within utility companies.

The ever-worsening mortality rates due to various diseases such as heart disease, breast cancer, and kidney disease are of great concern. Early diagnosis of the disease can be of great help. This process can be automated with the help of Artificial intelligence (AI). But, the main worry of using AI in healthcare is its black-box behaviour. The majority of the models characterized by high accuracy are often black-box in nature. This can be overcome by the use of eXplainable Artificial Intelligence (XAI), which is capable of explaining the predictions made by these black box models. We have exploited 3 different XAI frameworks: SHAP, LIME, and DALEX, to understand the working and the facilities provided by the three frameworks and compare them. We have used 5 disease datasets (3 heart disease, 1 cancer and 1 kidney disease) to carry out our work. Each dataset was trained with 3 machine learning models, namely Support Vector Machine (SVM), Logistic regression (LR), and K-Nearest neighbours (KNN), and the best model was used to feed to the XAI framework. LR performed best for one of the heart disease datasets with 72.31%accuracy, while SVM outperformed in all the other datasets, thus proving the efficacy of such approaches for early disease prediction.

The implementation of e-government initiatives remains a complex socio-technical challenge, particularly for administrations lacking structured knowledge of the Critical Success Factors (CSF) of E-government. The CSFs refer to the essential elements that must be effectively addressed to ensure the achievement of organizational objectives. In the context of E-government, CSFs encompass key determinants such as leadership commitment, technological infrastructure, user trust, policy support, and citizen engagement that collectively drive successful digital governance implementation. Governments often struggle to operationalize strategies and allocate resources effectively due to the absence of empirically grounded frameworks. To fill these gaps, this study combines systematic evidence synthesis, qualitative factor clustering, and quantitative multi-criteria validation to model and prioritize the interdependent CSFs governing sustainable e-government implementation across technological, organizational, human, environmental, and governance dimensions. The research employs a three-phase methodological pipeline: (1) Systematic Literature Review (SLR) guided by Preferred Reporting Items of Systematic Reviews and Meta-Analyses (PRISMA) standards to identify CSFs across more than 58 peer-reviewed studies; (2) Thematic Coding and Factor Clustering using NVivo-based qualitative content analysis to categorize determinants into organizational, technological, environmental, human, and governance domains; and (3) Analytic Hierarchy Process (AHP) Validation to assign relative weightings and interdependencies among identified factors. A total of 62 CSFs were extracted and classified under eight major domains: Strategic Planning and Governance, Planning and Execution Efficiency, Technical and Operational Aspects, User-Centric Focus and Quality Assurance, Technological Factors, Organizational Factors, Socio-Political Factors, and Economic Factors. Among these, User-Centric Focus and Quality Assurance (C₄) emerged as the most influential cluster with the highest global weight of 0.286, reflecting the growing emphasis on citizen trust, service quality, and satisfaction in digital governance systems. The top three CSFs identified through AHP were “Building Trust with Users” (LW = 0.266, GW = 0.033), “Visionary Leadership” (LW = 0.265, GW = 0.040), and “Comprehensive Planning” (LW = 0.275, GW = 0.038), representing the intersection of governance, user engagement, and execution excellence. This study contributes a decision-support framework that integrates both quantitative prioritization and qualitative contextualization, serving as a practical tool for policymakers, digital transformation officers, and public sector reform strategists.

This study investigates the fusion of artificial intelligence (AI) and group dynamics by examining undergraduate student perceptions (n=112) of AI tools (e.g., ChatGPT, Grok, Gemini, Grammarly, etc.) in collaborative group work at IMC Krems of Applied Sciences University, Tashkent campus. By using surveys, thematic analysis, it explores AI impact on communication, equity, and task management in culturally diverse, multilingual settings. Results show majority students regularly use AI tools for idea generation, feedback, language support. Qualitative analysis reveals four themes: enhanced efficiency, improved communication support and concerns about over-reliance and reduced interpersonal interaction. While AI serves as cognitive and emotional scaffolding but requires mindful, ethical integration to maximize benefits. The research offers novel insights for non-Western multilingual contexts and practical guidance for educators implementing human-AI hybrid collaboration.

Coronary heart disease, a prevalent cardiovascular condition, affects coronary arteries, causing progression over time. Factors include diabetes, hypertension, inactivity, and tobacco use. Treatment includes medications and surgery, while maintaining a balanced diet and regular physical activity can prevent it. This research aimed to develop and validate a predictive model for CHD occurrence, leveraging the power of multiple regression while considering a range of predisposing variables. This study uses a quantitative, retrospective design utilizing multiple regression analysis to predict the likelihood of coronary heart disease (CHD). The study involved 130 participants aged 24-85, with health history data on cardiovascular risk factors, blood pressure, cholesterol, smoking, BMI, and family history of heart disease. Multiple regression analysis was utilized to determine the significant predictors of CHD diagnosis. Significant relationships between responder variables and predictor factors in a multiple linear function are identified using multiple regression analysis. Our model discovered that a higher risk of coronary heart disease (CHD) was closely associated with both total cholesterol and BMI. The model included factors like systolic blood pressure, diabetes, physical activity, and smoking, but they had lower contributions to the prediction equation, despite cholesterol and BMI being the best predictors. This study successfully developed a multiple regression-based prediction model for CHD that can contribute to a more informed and potentially proactive approach to cardiac healthcare. Further work should focus on refining model accuracy and real-world clinical application.

In the field of computer vision, face recognition is a critical research area that has many applications in different fields such as security and medical treatment to authentication systems. Tradition feature descriptors are popular, but they are often handicapped by problems such as changes in lighting, posture and facial expression. While these techniques encode certain features well, they are subject to a number of biases including light sensitivity and computational complexity. In this paper, we present a new feature descriptor, the Directional Intensity Pattern (DIP) descriptor. It is an excellent combination of local texture, gradient magnitude and direction features. Feature selection and dimensionality reduction: Principal Component Analysis (PCA) for dimension reduction to improve discriminative power and less redundancy The Least Absolute Shrinkage and Selection Operator (LASSO) is used for feature selection. Furthermore, pre-processing techniques such as gamma correction and contrast normalization improved lightness invariance, thus increasing recognition performance. In this work, the DIP descriptor was evaluated on two public available datasets (YaleB, Face96). The results showed that it could achieve 97.59% and 98.36% accuracy on these datasets respectively, higher than the state-of-the-art methods. The result confirmed DIP descriptor remarkable ability to grasp quite a few texture and structure features of the picture in this manner it provides a powerful framework for face recognition under various circumstances.

The advancement and expansion of artificial intelligence (AI) has revolutionized traditional education paradigms. The ability of language models to process human language has revolutionized the field of artificial intelligence. This had led to the integration of language models such as Generative AI (GAI) into learning as it can understand and process human language efficiently. Fusion of these models has significantly enhanced education and research development leading to academic progress. There is gap in the learning needs of the students. Traditional teaching methods often fail to provide personalised adaptive environments and hence to fill this gap this research focusses on integration of AI tools in classrooms.  The objective of this paper is to explore and analyze the applications of integration of generative AI strategies in teaching and to examine the impact from educators’ perspective. The objective of the study is to evaluate the effectiveness of GAI powered integration in teaching and learning by analyzing the feedback scores gathered by students and teachers of an undergraduate course. Data was collected and analyzed using standard mean comparisons. Results of the analysis demonstrate that generative AI assisted teaching facilitated adaptive learning, automated content generation, enhanced student engagement and the quality of dynamic learning when compared with conventional strategies. Using quantitative analysis, the study validates GAI fusion, and the data is analyzed using standard mean scores.  The improvement performance of students and educators feedback for traditional and GAI is 56.63% and 54.41% respectively, which suggests a positive shift of moving from traditional to GAI approaches. This strong score shows the GAI approach is more effective and student-centered. The results reveal that though challenges exist, strategic guided integration of GAI significantly enhances pedagogical factors of education and thus plays a crucial role in shaping AI education as AI models evolve.

Surveillance cameras play a pivotal role in educational institutions. They monitor the educational process, detect violations, and protect students from potential injuries or dangers. Continuous recording generates a massive amount of video data. Human observers spend significant time and effort reviewing the footage. Reviewing aims to detect and quickly address abnormal events. Abnormal events are rare in educational environments. Observers may become bored during continuous monitoring. This may cause fatigue and loss of attention. To overcome these challenges, this paper proposes an intelligent system that combines summarization and abnormal event detection in surveillance video. It is divided into two stages: The first stage starts with the extraction of static, feature-based key frames that highlight the video's most significant content. In the second stage, Convolutional Autoencoder (CAE) network used to detect abnormal events from the key frames generated by the summary stage. The proposed system produces two separate videos: a general summary and a dedicated abnormal events video sent to the relevant individuals. The proposed system was tested on some benchmark datasets. The experimental results demonstrated that the proposed system was effective in reducing browsing time and effort, as well as in detecting abnormal events within an educational context.

Advancements in high-speed communication networks, such as 5G and 6G, display the shortcomings of earlier Radio Frequency (RF) systems due to their limited access to the electromagnetic spectrum. Optical Wireless Communication (OWC) gives access to an unlimited optical spectrum that can address the demands in 6G networks. One key thing about Free Space Optical (FSO) is that it uses the near-infrared spectrum to transfer large amounts of data over several kilometers. FSO systems can be found in a large number of places, ranging from home and outdoor use to important roles in the military and in medical settings. These systems, however, struggle to transmit signals clearly and reliably when the distance is very long due to effects of the atmosphere. One solution to these problems is to rely on advanced channel modeling and using Multiple-Input Multiple-Output (MIMO) schemes, as they improve reliability and efficiency. The latest research efforts are centered on Massive MIMO-FSO networks that make use of spatial diversity to fight atmospheric fading and guarantee a sturdier connection. Importantly, Machine Learning (ML) is transforming the way research is carried out. Channel estimation, turbulence prediction, signal demodulation, and adaptive modulation can now be done using ML, which reduces the need for many calculations and makes things run more smoothly. Using information from data, ML helps optimize FSO systems in different channel conditions. This study provides a review of how machine learning is applied in Massive MIMO-FSO systems. It sorts out highlighting current strategies, explaining their strengths, weaknesses, and how to use them. The main goal of this review is to give an in-depth look at how ML-assisted optical wireless systems can fulfill the needs of future communication networks.

This paper presents a novel machine-learning framework designed to personalize Cognitive Behavioral Therapy (CBT) for adult patients by leveraging a multi-dimensional, adaptive approach. The proposed system integrates historical clinical data, real-time behavioral indicators, and contextual factors to generate a comprehensive psychological profile for each adult patient. A reinforcement learning mechanism underpins therapy selection, allowing the model to iteratively refine treatment strategies based on individual responses and therapeutic outcomes. An embedded optimization process enables dynamic adaptation of interventions, improving predictive accuracy and fostering patient-centered care. The framework incorporates a multi-factor assessment model that synthesizes psychological, behavioral, and physiological variables to enhance therapeutic effectiveness, sustainability, and responsiveness to change. Comparative evaluations demonstrate that this approach outperforms traditional CBT planning methods, as well as existing deep learning, hybrid, and reinforcement-based models, in terms of accuracy, interpretability, computational efficiency, and patient outcome optimization for adults. Furthermore, the system emphasizes fairness and equity in treatment personalization, supporting real-time clinical decision-making while minimizing ineffective therapeutic pathways. This research underscores the transformative potential of machine learning in mental health care by enabling scalable, data-driven, and continuously improving interventions tailored to the nuanced needs of adult patients undergoing CBT.

This research introduces a cutting-edge deep learning-based agricultural engineering illness diagnosis approach. Convolutional neural networks (CNNs) and improved methods improve accuracy and efficiency. The recommended solution includes network settings, convolution processes, and sharing strategies to reduce dimensions. These methods reduce the network's processing power so it can concentrate on disease characteristics. The model employs dropout regularization, attention processes, and multi-scale feature extraction to enhance sickness prediction. The technology also utilizes photographs and sensor data to adapt to agricultural circumstances. The performance test shows that the suggested technique outperforms traditional machine learning and mixed models in F1 score (95%), accuracy (95%), precision (94%), memory (96%), and correctness (94%). It has high discriminative power with an AUC-ROC score of 0.98. The model uses computers well: two hours to train, two seconds to derive conclusions, and 65% of the CPU at all times. Real-time farming could benefit from its use. The suggested technique can properly and reliably diagnose illnesses due to its low overfitting rate and excellent generalization potential. The precision agriculture technique will enhance crop health management and productivity.

Social media has supplanted conventional media as one of the most important venues for information exchange. Because of the internet's accessibility and simplicity, news on community media tends to spread quicker and simpler than a conventional news source. Still, not all of the information shared on ‘social media’ is true and/or comes from untrustworthy sources. Fake news may readily be manufactured and disseminated throughout ‘social media’, and this counterfeit news has the potential to mislead or misinform readers. Though several physical fact-inspection websites have been built to determine if the news is reliable, they cannot keep up with the amount of rapidly circulated internet information, particularly on social media. Twitter, being one of the most well-known continuing news sources, also happens to be one of the most dominating news disseminating media. Topic models facilitate the detection of the most relevant vocabulary and concept within a text corpus. This paper proposes a model for recognizing fake news messages from twitter posts using a novel data-mining algorithm. Here initially the twitter dataset is collected preprocessing is done by using word embedding. ‘Term Frequency Inverse Document Frequency ‘(TF-IDF)’ and Latent Semantic Analysis (LSA) do feature extraction. Feature selection is based on the Adaptive Whale Optimized Wrapper (AWOW) method. We proposed Fine-tuned Weighted Probabilistic Bayesian Neural Network (FWP-BNN) for the classification of the normal and the fake news. The proposed method is compared with existing approaches and the metrics are evaluated. The efficacy of the suggested technique in recognizing fake tweets is shown by test findings on a large miscellaneous events dataset.

Chronic diseases (CDs) have become as significant as communicable diseases due to their rising mortality rates and long-term effects. Coronary artery disease (CAD), one of the most common NCDs, is increasingly concerning due to its impact on both death rates and overall health. Managing CAD typically requires professional care and lifestyle changes, which may be inaccessible to some patients due to financial constraints or difficulty in modifying their habits. However, remote health solutions, like mobile applications, could help CAD patients improve their condition and lower risks. In Iraq, the willingness of CAD patients to use mHealth apps has not been explored. This study examines existing mHealth readiness models and incorporates additional factors that consider the needs of CAD patients and the Iraqi context. This will be achieved by adapting a questionnaire based on expert feedback and distributing it to CAD patients in Iraq.