Search

Abstract Enhancement of security, personalization, and safety in advanced transportation systems depends on driver identification. In this context, this work suggests a new method to find drivers by means of a Random Forest model optimized using the osprey optimization algorithm (OOA) for feature selection and the salp swarm optimization (SSO) for hyperparameter tuning based on driving behavior. The proposed model achieves an accuracy of 92%, a precision of 91%, a recall of 93%, and an F1-score of 92%, significantly outperforming traditional machine learning models such as XGBoost, CatBoost, and Support Vector Machines. These findings show how strong and successful our improved method is in precisely spotting drivers, thereby providing a useful instrument for safe and quick transportation systems.

The increasing use of Automatic Dependent Surveillance-Broadcast (ADS-B) technology in flight control systems has created many serious concerns.These weaknesses threaten the security and safety of our aviation industry. Therefore, to enhance aviation control security and better deal with these problems, this research focuses on developing a strong ADS-B injection detection system. It combines XGBoost and Random Forest with Logistic Regression in an Ensemble Learning Meta-Learning Model to identify ADS-B injection risks and categorise them. Ensemble methods, which combine several models can increase the detection accuracy and robustness of the model used to identify the threat. In addition, Explainable AI (XAI) methods are employed to enhance the process of explaining how the model reaches its decisions and building trust in aviation security systems. The system’s training, testing, and evaluation are conducted with ADS-B data. This result indicates that the Stacked Random Forest and XGBoost with Logistic Regression Meta-Learner with 99.60% accuracy, along with good recall rates (99.49%) and precision (99.41%). Also, aviation control authorities are reassured by the model’s transparent and applicable decision logic through the application of XAI techniques. This research contributes to enhanced aviation security by proposing a new, highly accurate ADS-B injection detection system with explainable outcomes. A strategy like this can help flight control systems maintain integrity amidst an ever-digitising aviation reality.

Abstract Drought is a natural disaster that can affect a larger area over time. Damage caused by the drought can only be reduced through its accurate prediction. In this context, we proposed a hybrid stacked model for rainfall prediction, which is crucial for effective drought forecasting and management. In the first layer of stacked models, Bi-directional LSTM is used to extract the features, and then in the second layer, the LSTM model will make the predictions. The model captures complex temporal dependencies by processing multivariate time series data in both forward and backward directions using bi-directional LSTM layers. Trained with the Mean Squared Error loss and Adam optimizer, the model demonstrates improved forecasting accuracy, offering significant potential for proactive drought management.

This paper provides a bibliometric overview of the interaction between environmental education and the growth of sustainable entrepreneurship. Using VOSViewer and the R programming language, we found 849 reference studies for investigation while browsing English-language articles from esteemed publications and conferences recorded in the Scopus databases. Our analysis revealed an exponential increase in scholarly publications on the topic, therefore highlighting its development as a current focus of study. Our bibliometric study reveals a major knowledge gaps despite this increase: knowledge of the successful integration strategies of environmental education into sustainable entrepreneurship practices and policies at different levels (local, national, and global). Moreover, thorough research on how these interactions affect environmental awareness across several populations and the accomplishment of sustainable development goals (SDGs) is lacking. Therefore, our study aims to close this important void by offering a thorough review of the present situation of research, pointing out the best journals, countries, and authors leading in this field, and suggesting future directions to improve the synergy between environmental education and sustainable development. Policies, courses, and projects aiming at a more sustainable future depend on this input in setting their direction.

The early detection of brain tumor is crucial for effective treatment and improved patient prognosis in Industrial Information Systems. This research introduces a novel computational model employing a three-layer Convolutional Neural Network (CNN) for the identification of brain tumors in Industrial Information Systems. Leveraging advanced computational techniques, this proposed model can autonomously detect intricate patterns and features from medical imaging data, resulting in more accurate and expedited diagnoses. With an impressive 90 % precision rate, our model demonstrates the potential to serve as a valuable tool for medical professionals working in the field of neuroimaging. By presenting a dependable and precise computational model, this study contributes to the advancement of brain tumor identification within the domain of medical imaging. We anticipate that our methodology will aid healthcare providers in making more accurate diagnoses, thereby leading to enhanced patient outcomes. Potential avenues for future research encompass refining the model's fundamental architecture and exploring real-time therapeutic applications.

As Intelligent Transport Systems (ITS) continue to evolve, the quest for improving road safety and transportation efficiency has gained renewed emphasis. One of the pivotal aspects in this endeavor is the detection and analysis of driver behavior. Recognizing signs of fatigue, distraction, or inattentiveness is critical in enhancing road safety and optimizing traffic flow. In this paper, we present a pioneering approach to driver behavior detection within the realm of ITS using deep learning models in the Cyber-Physical Systems (CPS) framework. Our research focuses on the discernment of critical behaviors such as eye closure, open-eye state, yawning, and non-yawning instances. With an unwavering commitment to road safety and transportation efficiency, we’ve harnessed the power of deep learning to design, develop, and train an exceptionally accurate model. Through rigorous evaluation, we achieved an impressive 94% accuracy. Our findings unveil the potential of CPS-based solutions for real-time driver behavior monitoring, providing a foundation for safer roadways and more streamlined traffic management. The proposed deep learning model offers robust and accurate predictions, enabling timely responses to various driving conditions. This research significantly advances the field of driver behavior analysis within the context of intelligent transportation systems, with broad implications for road safety and traffic management.

In today's world, where digital threats are on the rise, one particularly concerning threat is the Mirai botnet. This malware is designed to infect and command a collection of Internet of Things (IoT) devices. The use of Mirai attacks has intensified in recent times, thus threatening the smooth operation of numerous devices that are connected to a network. Such attacks carry adverse consequences that include interference with services or the leakage of confidential information. To fight this growing threat, smart and flexible detection techniques are required to counter the new methods cyber attackers use. The aim of this research is to develop a resilient defense against Mirai botnet attacks. The Long Short Term Memory term (LSTM) and XGBoost combined have the best performance of 97.7% accuracy score. With this combination, the aim is to strengthen our cyber defenses against sophisticated and dynamically operating Mirai botnets to further enhance the security of our digital world.

Sustainable entrepreneurship refers to the practice of creating and managing a business venture in a manner that is environmentally and socially responsible while still maintaining financial profitability. Social media has become essential for entrepreneurs to build and promote their businesses. It allows them to connect with potential customers, build brand awareness, and reach a wider audience. In recent years, there has been increasing interest in exploring the relationship between sustainable entrepreneurship and social media and how social media can impact sustainable entrepreneurship development. This paper aims to analyze social media's impact on sustainable entrepreneurship development. The study uses data collected from Scopus to identify keywords related to social media, entrepreneurship, and sustainability. Coupling analysis identifies the most frequent keyword groups and clusters and measures their centrality, impact, and color. The results show that social media and entrepreneurship are strongly linked and that social media can positively impact sustainable entrepreneurship development. The study also highlights the importance of ecosystem thinking in promoting sustainable entrepreneurship and the potential of machine learning and sentiment analysis in studying the impact of social media on entrepreneurship. The findings of this study can be useful for policymakers, entrepreneurs, and researchers interested in promoting sustainable entrepreneurship through social media.

Phishing is a cybersecurity problem that hackers employ to deceive individuals and organizations. Phishing is dynamic in nature; the hackers change several tricks to deceive the victims in multiple ways. It is important to track the tricks of hackers with recent technology. This study makes a notable contribution to enhancing cybersecurity defences by offering insights that aid in the detection and mitigation of phishing threats. Specifically, the study’s analysis of URLs using mutual information and logistic regression techniques yielded a remarkably high accuracy rate of 99.97%, surpassing previous efforts. The identification of the most informative features for distinguishing phishing attempts provides valuable intelligence for cybersecurity professionals, enabling them to bolster defenses and stay ahead of evolving phishing tactics.

Internet of Things (IoT) devices within smart cities, require innovative detection methods. This paper addresses this critical challenge by introducing a deep learning-based approach for the detection of network traffic attacks in IoT ecosystems. Leveraging the Kaggle dataset, our model integrates Convolutional Neural Networks (CNNs) and Gated Recurrent Units (GRUs) to capture both spatial and sequential features in network traffic data. We trained and evaluated our model over ten epochs, achieving an impressive overall accuracy rate of 99%. The classification report reveals the model’s proficiency in distinguishing various attack categories, including ‘Normal’, ‘DoS’ (Denial of Service), ‘Probe’, ‘U2R’ (User to Root), and ‘Sybil’. Additionally, the confusion matrix offers valuable insights into the model’s performance across these attack types. In terms of overall accuracy, our model achieves an impressive accuracy rate of 99% across all attack categories. The weighted- average F1-score is also 99%, showcasing the model’s robust performance in classifying network traffic attacks in IoT devices for smart cities. This advanced architecture exhibits the potential to fortify IoT device security in the complex landscape of smart cities, effectively contributing to the safeguarding of critical infrastructure

MANET is a self-organized and infrastructure-free network. Therefore, the use of Virtual Backbone Network is becoming popular nowadays because it provides stability to MANETs. In virtual backbone network only selected nodes stores the topological information about the network. So, only the selected nodes are involved in the routing of the messages. In past recent years, many virtual backbone construction algorithms have been proposed by the researchers but not much work has been done in the security of virtual backbone network. In this paper, we discuss this issue by proposing a secure lightweight backbone construction approach. We have used a lightweight distributed approach for securing the backbone network, due to which our proposed approach is capable to detect many existing attacks without draining out the resources of the nodes. We have used the NS-2 simulator to check the efficiency of our proposed approach. Simulation results show that implementing our proposed approach does not drain the resources of the nodes and it also retains a high degree of detection rate without bringing in any significant traffic.

The pervasive integration of the Cognitive Internet of Things (CIoT) in advanced communication systems has significantly heightened security concerns due to inherent vulnerabilities that cyber attackers exploit. These security challenges are exacerbated by IoT devices’ limited computational resources and storage capacities, rendering traditional security protocols ineffective. This paper introduces a AI- based security model tailored explicitly for IoT environments within advanced communication systems. Our approach leverages AI techniques for cyber attack detection and mitigation. The effectiveness of our model is demonstrated through its capability to identify different types of cyberattacks with a precision of more than 99%. This study advances the security mechanisms for IoT devices and offers a scalable and efficient solution to safeguard user privacy and system integrity in sophisticated communication networks.

In current scenario when economic and risk management sectors need accurate predictions of enterprise bankruptcy, it is very importance issue to research in the field of security of enterprise bankruptcy. In this context, we propose an hybrid deep learning model through the use of convolutional neural network to enhance bankruptcy forecasting models. We address the high-dimensional data and imbalanced problems by introducing feature selection strategically and Synthetic Minority Over-sampling Technique (SMOTE). In a comparative evaluation, the performance of our model is over 81 %, which is better than that for Logistic Regression and Support Vector Machines. This leap in accuracy demonstrates the cutting edge unprecedented ability of our model to decrypt complex financial patterns and establishes a new precedent for deep learning applications in the nuanced field of financial analytics.

The rise in mobile malware risks brought on by the explosion of Android smartphones required more efficient detection techniques. Inspired by a cascade of Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks, optimized using the Earthworm Optimization Algorithm (EOA), this study presents an android malware detection model. The paper used random forest model for feature selection. With a 99% accuracy and the lowest loss values, the proposed model performs better than conventional models including GRU, LSTM, RNN, Logistic Regression, and SVM.. The findings highlight the possibility of proposed method in improving Android malware detection, thereby providing a strong answer in the changing scene of cybersecurity.