Traffic management is a critical task in software-defined IoT networks (SDN-IoTs) to efficiently manage network resources and ensure Quality of Service (QoS) for end-users. However, traditional traffic management approaches based on queuing theory or static policies may not be effective due to the dynamic and unpredictable nature of network traffic. In this paper, we propose a novel approach that leverages Graph Neural Networks (GNNs) and multi-arm bandit algorithms to dynamically optimize traffic management policies based on real-time network traffic patterns. Specifically, our approach uses a GNN model to learn and predict network traffic patterns and a multi-arm bandit algorithm to optimize traffic management policies based on these predictions. We evaluate the proposed approach on three different datasets, including a simulated corporate network (KDD Cup 1999), a collection of network traffic traces (CAIDA), and a simulated network environment with both normal and malicious traffic (NSL-KDD). The results demonstrate that our approach outperforms other state-of-the-art traffic management methods, achieving higher throughput, lower packet loss, and lower delay, while effectively detecting anomalous traffic patterns. The proposed approach offers a promising solution to traffic management in SDNs, enabling efficient resource management and QoS assurance.
5 January 2025 in Journal of Machine and Computing
Identifying and classifying diabetes problems among women can be achieved using several Machine Learning (ML) algorithms. This paper additionally includes a summary of the evaluation of the performance of these MLs with algorithms on many different classification metrics. The AUC-ROC score is the best for Extreme Gradient Boost (XGB) with 85%, followed by SVM and Decision Trees (DT). Logistic Regression (LR) is showing low performance. However, the DT and XGB show promising performance against all the classification metrics. However, the SVM shows a lower support value; hence, it cannot be claimed to be a precious classifier. A study reveals that women are four times more susceptible to diabetic conditions than men. But the healthcare systems do not give special attention to diabetic conditions in women. This study proposes to predict the probability of diabetes in females based on numerous medical conditions they may have. The ML accurately predicts diabetic complications based on biological conditions such as blood glucose levels, age, Body Mass Index (BMI), numerous pregnant women, and other factors.
Traffic shaping is a critical task in software-defined -IoT networks (SDN-IoTs) to efficiently manage network resources and ensure Quality of Service (QoS) for end-users. However, traditional traffic shaping approaches based on queuing theory or static policies may not be effective due to the dynamic and unpredictable nature of network traffic. In this paper, we propose a novel approach that leverages Graph Neural Networks (GNNs) and Multi-arm Bandit algorithms to dynamically optimize traffic shaping policies based on real-time network traffic patterns. Specifically, our approach uses a GNN model to learn and predict network traffic patterns and a Multi-arm Bandit algorithm to optimize traffic shaping policies based on these predictions. We evaluate the proposed approach on three different datasets, including a simulated corporate network (KDD Cup 1999), a collection of network traffic traces (CAIDA), and a simulated network environment with both normal and malicious traffic (NSL-KDD). The results demonstrate that our approach outperforms other state-of-the-art traffic shaping methods, achieving higher throughput, lower packet loss, and lower delay, while effectively detecting anomalous traffic patterns. The proposed approach offers a promising solution to traffic shaping in SDNs, enabling efficient resource management and QoS assurance
AI-powered autonomous vehicles (AVs) have led to discussions about who is responsible and answerable for problems that arise in human-AI collaborations. The research focuses on how the decisions of machine learning algorithms in AVs are interpreted by the law. With the use of Decision Trees, Random Forests, Support Vector Machines and Deep Neural Networks, this research examined how driver behavior is predicted in AVs across a controlled set of scenarios. Of all the algorithms, the Deep Neural Network showed the best accuracy at 94.5%, then Random Forest at 91.2% and the other two fell below at 87.6% and 83.4%. Yet, accuracy was negatively linked to interpretability and Decision Trees were the model that offered the easiest trace of logic. Learning from both comparative experiments and recent law and ethics writings, the discipline points out that the law is currently unable to handle situations where a decision is made independently by a non-human agent. The results suggest putting in place a model in which developers, manufacturers and AI systems are all held accountable. The research introduces a system that can guide policymakers in writing future regulations for AI technology in mobility.