Brain tumors are deemed accurately diagnosed by MRI imaging for prompt treatment. This study presents a federated learning-enhanced edge network brain tumor prediction architecture that addresses data privacy, latency, and computational efficiency issues. The framework analyzes data locally using edge computing, lowering latency and providing real-time diagnostics. Federated learning keeps confidential patient data on local devices while enabling collaborative model training over edge nodes. To demonstrate the pros and cons of decentralized machine learning, federated learning models are bench marked against centralized methods. This research shows that federated learning on edge networks revolutionize medical diagnostics by delivering a robust, scalable, and privacy-preserving MRI brain tumor prediction solution.

The proposed investigation implies a state-of-theart design of a real-time system, which is optimized by AI, adopts AMQP, NNPC, and DES. The Advanced Message Queuing Protocol ensures not only low latency but also reliability in the communication process, Neural Network Predictive Control enables us to adjust predictive control according to the changes in the nonlinear environment, and the Discrete Event Simulation enables us to perform performance analysis and optimise decisions based on the simulation. The suggested framework is experimented with the help of discrete-event simulation, which implements the simulation of event flow and system dynamics in real time. The comparison and evaluation of the performance with the traditional ones, which include Agent-driven modelling, hybrid prediction management, and Constrained Application Protocol communications, is conducted in terms of model performance and evaluation criteria. According to the findings of the simulation, it is possible to provide high performance proposed Advanced Message Queuing Protocol - Neural Network Predictive Control - Discrete Event Simulation system demonstrates its efficiency and effectiveness even in real-time operational conditions.

Climate-sensitive illnesses like malaria and climate-related fever largely rely on climatic factors like temperature, humidity, and rainfall. Classical models are unlikely to generalize over different geographies and climatic conditions and need more interpretable and robust solutions. In this paper, a new AI predictive model has been introduced by grounding it in a stacking ensemble of regressors enriched with explainable AI (XAI) methods which basically aggregates many base learners into a meta-learner and uses SHAP and LIME to effect global and local interpretability of predictions. Tested on 1,456 climate-health instances, the proposed model outperformed individual models (R2 = 0.9489, MAE = 7.41), effecting better predictive accuracy and transparent decision reasoning, making it more reliable and actionable for practitioners.

Infectious/contagious diseases remain a significant global health challenge, necessitating accurate identification to mitigate their spread. The widespread adoption of wearable healthcare devices capable of continuously monitoring physiological parameters presents a unique opportunity for enhancing disease detection strategies. In this article, we develop a new data fusion-enabled explainable artificial intelligence-assisted light gradient boosting model (LightGBM) (FuXAI) model in edge networks to predict contagious disease in the early stage. The overall contributions of the proposed FuXAI model are threefold. First, we create a comprehensive health profile for each individual by employing data fusion techniques and integrating health parameters received from multiple sources. Second, we integrate a new lightweight machine learning (ML) model, the LightGBM, to predict the disease at resource-constraint edge devices over the fused data. Finally, we leverage the power of explainable artificial intelligence (XAI) to develop interpretable algorithms, integrating with the LightGBM that can identify subtle patterns and correlations within the fused data, potentially revealing early warning signs of infectious diseases and creating trust in medical professionals during decision-making. Extensive simulation results of the proposed model over the standard ML models using benchmark datasets demonstrate the effectiveness of the model.

Edge computing minimizes latency and enhances real-time data processing compared to traditional cloud systems. However, efficient resource management and task scheduling remain complex due to dynamic workloads, heterogeneous nodes, and limited energy budgets. This study introduces a hybrid metaheuristic optimization and graph-based scheduling framework for adaptive resource management in edge environments. The proposed system integrates quasi-Sobol optimization with sigmoid fuzzy logic to balance computational load and minimize latency, while directed acyclic graph (DAG)-based scheduling ensures dependency-aware execution across edge nodes. Adaptive optimization control strengthens scalability and responsiveness under varying workload conditions. The framework is validated through simulation using edge server datasets and evaluated based on precision, recall, F1-score, latency, and energy consumption metrics. Experimental results demonstrate significant improvements in task allocation accuracy, resource utilization, and power efficiency compared to conventional scheduling methods. The approach offers a technically robust and scalable foundation for real-time, resource-efficient task scheduling in modern edge computing applications.

Communicable diseases are transmitted through water, food, contaminated surfaces, bodily fluids, air. In such a situation, staying in home isolation for fewer chronic health problems and monitoring health status frequently through Medical Sensors (MSs) is recommended. The use of Artificial Intelligence (AI) in smart consumer electronics and sustainable healthcare has recently demonstrated remarkable results. However, the healthcare domain requires high levels of accountability and transparency for communicable disease prediction and sustainable life in edge networks. This paper aims to present an intelligent healthcare prototype that can identify risk factors according to monitoring parameters by analyzing the Explainable XGBoost (XXGB) model. Using edge networks for sustainable living, we explore the intersection between healthcare and consumer electronics. Initially, the prototype has been trained using the XXGB model over one publicly available dataset related to communicable diseases. Next, the prototype identifies patient risk factors by analyzing real-time monitoring parameters. Simulation results illustrate the efficiency of the proposed XXGB model up to 84.2% accuracy, which is higher than existing models.

Vector-borne diseases have created critical world-wide public health hazards that require innovative approaches for prevention and management. Within the context of IoT-enabled edge networks, this study offers a unique framework to build a reliable and straightforward distant prediction model for vector-borne diseases. The suggested strategy incorporates the use of edge computing capabilities to predict the onset and spread of diseases caused by vectors by using real-time data streams from mobile applications. Employing widely used techniques from machine learning, including interpretable models, our system not only achieves high prediction accuracy but also provides insight into the most important factors influencing the spread of disease. Moreover, this approach overcomes the resource constraints essential to edge networks, ensuring the prediction model's dependability and scalability.

Due to the expanding growth of populations worldwide, infectious disease expenses are increasing rapidly. Patients with lesser chronic health problems are often encouraged to recover at home, which requires continuous remote monitoring and recommendation systems for decision-making. To overcome the above-mentioned challenges and reduce connectivity between infected patients and medical officers, we develop an artificial intelligence (AI)-based healthcare prototype in edge networks to monitor and analyze health parameters remotely. In the first phase of the work, health parameters are collected through multiple biosensors, including temperature, pulse rate, and oxygen saturation, and stored at the edge of the network. In the second phase, the monitoring data are preprocessed and analyzed using the neural network (NN) model in the edge devices for decision-making about the health status of the infected patient. Finally, the performance of the proposed model is evaluated using a real-time dataset, and multiple performance metrics are collected from the developed prototype to assess its effectiveness. The extensive simulation results demonstrate the efficiency of the proposed model over the existing ML models, i.e., the NN model achieves 94.05% accuracy, which is higher than the existing ones.

The imbalanced dataset's existing classification methods have low prediction accuracy for the minority class because of the little information present. Using over- and under-sampling techniques, we can improve the minority's ability to forecast outcomes. However, the minority class's accuracy of prediction is negatively impacted by the two methods due to the loss of vital information or the addition of irrelevant details for classification. SVM kernels have great abilities to handle asymmetric data, but when we need to use SVM kernels alone or as part of the ensemble technique for an unbalanced dataset, we don't have a strong reason to choose which kernel to use, and also how a particular kernel will act depends a lot on the data set. In this paper, we present a framework in which several kernel SVM (Linear, Polynomial, Sigmoid, RBF) classifiers were utilized as the base learners and one of the kernels (say RBF kernel) as meta learner using the Stacking Ensembles technique, which shows that stacked generalization of SVM kernels gives similar results as best performing kernel for an imbalanced dataset of software change proneness, using AUC, ROC, MCC, and BAS as an evaluation matrix.

The emergence of novel types of communication, such as email, has been brought on by the development of the internet, which radically concentrated the way in that individuals communicate socially and with one another. It is now establishing itself as a crucial aspect of the communication network which has been adopted by a variety of commercial enterprises such as retail outlets. So in this research paper, we have built a unique spam-detection methodology based on email-body sentiment analysis. The proposed hybrid model is put into practice and preprocessing the data, extracting the properties, and categorizing data are all steps in the process. To examine the emotive and sequential aspects of texts, we use word embedding and a bi-directional LSTM network. this model frequently shortens the training period, then utilizes the Convolution Layer to extract text features at a higher level for the BiLSTM network. Our model performs better than previous versions, with an accuracy rate of 97-98%. In addition, we show that our model beats not just some well-known machine learning classifiers but also cutting-edge methods for identifying spam communications, demonstrating its superiority on its own. Suggested Ensemble model's results are examined in terms of recall, accuracy, and precision

Nowadays, federated learning (F1) has been proposed as an emerging technology to store sensory data and train the edge networks using a set of computing devices with minimum training time. However, a collection of heterogeneous participating devices with different processing power and energy usage are used to analyze the model parameters locally. Therefore, an intelligent service provisioning mechanism with the F1 technique needs to be developed at the edge networks. This strategy can increase the security and privacy of the network while minimizing the training time on resource-constrained edge devices. In this magazine, we describe the importance of the FL-aided hybrid edge intelligent framework for next-generation Internet of Things applications. Moreover, to enhance the critical service provisioning functionality, we highlight two use case studies along with their potential research directions, including intelligent transportation systems and intelligent healthcare systems. Finally, this work concludes with a set of potential future research directions of FL-aided edge networks.