Search

Educational data mining has sparked a lot of attention in latest years. Many machine learning methods have been suggested to discover hidden information from educational data. The extracted knowledge assists institutions in enhancing the effectiveness of teaching tactics and the quality of education. As a result, it improves students' performance and educational outputs overall. In this paper, a classification model was built to classify students' grades in a specific course into different categories (binary and multi-level classification tasks). The dataset contains features related to academic and non-academic information. The models were built using a variety of machine learning algorithms: decision tree (J48), support vector machine (SVM), and k-nearest neighbor (K-NN). Furthermore, ensemble methods (bagging, boosting, random subspace, and random forest) which combined multiple decision tree classifiers were implemented to improve the models' performance. The data set was modified under two stages: features selection method and data augmentation using a method called synthetic minority over sampling technique (SMOTE). Based on the results of the experiments, it is possible to predict the students' performance successfully by using machine learning algorithms and ensemble methods. Random subspace obtained the best accuracy at two-level classification task with modified data with 91.20%. At the three-level classification task, the best accuracy was obtained by random forest with 87.18%. [ABSTRACT FROM AUTHOR]

The internet of things (IoT) represents a rapidly expanding sector within computing, facilitating the interconnection of myriad smart devices autonomously. However, the complex interplay of IoT systems and their interdisciplinary nature has presented novel security concerns (e.g. privacy risks, device vulnerabilities, Botnets). In response, there has been a growing reliance on machine learning and deep learning methodologies to transition from conventional connectivitycentric IoT security paradigms to intelligence-driven security frameworks. This paper undertakes a comprehensive comparative analysis of recent advancements in the creation of IoT botnets. It introduces a novel taxonomy of attacks structured around the attack life-cycle, aiming to enhance the understanding and mitigation of IoT botnet threats. Furthermore, the paper surveys contemporary techniques employed for early-stage detection of IoT botnets, with a primary emphasis on machine learning and deep learning approaches. This elucidates the current landscape of the issue, existing mitigation strategies, and potential avenues for future research. [ABSTRACT FROM AUTHOR]

This study presents a novel and innovative approach to automatically translating Arabic Sign Language (ATSL) into spoken Arabic. The proposed solution utilizes a deep learning-based classification approach and the transfer learning technique to retrain 12 image recognition models. The image-based translation method maps sign language gestures to corresponding letters or words using distance measures and classification as a machine learning technique. The results show that the proposed model ismore accurate and faster than traditional image-based models in classifyingArabiclanguage signs, with a translation accuracy of 93.7%. This research makes a significant contribution to the field of ATSL. It offers a practical solution for improving communication for individuals with special needs, such as the deaf and mute community. This work demonstrates the potential of deep learning techniques in translating sign language into natural language and highlights the importance of ATSL in facilitating communication for individuals with disabilities. [ABSTRACT FROM AUTHOR]

The exponential growth of Internet and network usage has necessitated heightened security measures to protect against data and network breaches. Intrusions, executed through network packets, pose a significant challenge for firewalls to detect and prevent due to the similarity between legitimate and intrusion traffic. The vast network traffic volume also complicates most networkmonitoring systems and algorithms. Several intrusion detection methods have been proposed, with machine learning techniques regarded as promising for dealing with these incidents. This study presents an Intrusion Detection System Based on Stacking Ensemble Learning base (Random Forest, Decision Tree, and k-Nearest-Neighbors). The proposed system employs pre-processing techniques to enhance classification efficiency and integrates sevenmachine learning algorithms. The stacking ensemble technique increases performance by incorporating three base models (Random Forest, Decision Tree, and k-Nearest-Neighbors) and a meta-model represented by the Logistic Regression algorithm. Evaluated using the UNSW-NB15 dataset, the proposed IDS gained an accuracy of 96.16% in the training phase and 97.95% in the testing phase, with precision of 97.78%, and 98.40% for taring and testing, respectively. The obtained results demonstrate improvements in other measurement criteria. [ABSTRACT FROM AUTHOR]

The smart city vision has driven the rapid development and advancement of interconnected technologies using the Internet of Things (IoT) and cyber-physical systems (CPS). In this paper, various aspects of IoT and CPS in recent years (from 2013 to May 2023) are surveyed. It first begins with industry standards which ensure cost-effective solutions and interoperability. With ever-growing big data, tremendous undiscovered knowledge can be mined to be transformed into useful applications. Machine learning algorithms are taking the lead to achieve various target applications with formulations such as classification, clustering, regression, prediction, and anomaly detection. Notably, attention has shifted from traditional machine learning algorithms to advanced algorithms, including deep learning, transfer learning, and data generation algorithms, to provide more accurate models. In recent years, there has been an increasing need for advanced security techniques and defense strategies to detect and prevent the IoT and CPS from being attacked. Research challenges and future directions are summarized. We hope that more researchers can conduct more studies on the IoT and on CPS. [ABSTRACT FROM AUTHOR]

Social media networks are becoming essential to our daily activities, and many issues are due to this great involvement in our lives. Cyberbullying is a social media network issue, a global crisis affecting the victims and society as a whole. It results from a misunderstanding regarding freedom of speech. In this work, we proposed a methodology for detecting such behaviors (bullying, harassment, and hate-related texts) using supervised machine learning algorithms (SVM, Naïve Bayes, Logistic regression, and random forest) and for predicting a topic associated with these text data using unsupervised natural language processing, such as latent Dirichlet allocation. In addition, we used accuracy, precision, recall, and F1 score to assess prior classifiers. Results show that the use of logistic regression, support vector machine, random forest model, and Naïve Bayes has 95%, 94.97%, 94.66%, and 93.1% accuracy, respectively. [ABSTRACT FROM AUTHOR]

Lung cancer has been the leading cause of cancer death for many decades. With the advent of artificial intelligence, various machine learning models have been proposed for lung cancer detection (LCD). Typically, challenges in building an accurate LCD model are the small-scale datasets, the poor generalizability to detect unseen data, and the selection of useful source domains and prioritization of multiple source domains for transfer learning. In this paper, a multiround transfer learning and modified generative adversarial network (MTL-MGAN) algorithm is proposed for LCD. The MTL transfers the knowledge between the prioritized source domains and target domain to get rid of exhaust search of datasets prioritization among multiple datasets, maximizing the transferability with a multiround transfer learning process, and avoiding negative transfer via customization of loss functions in the aspects of domain, instance, and feature. In regard to the MGAN, it not only generates additional training data but also creates intermediate domains to bridge the gap between the source domains and target domains. 10 benchmark datasets are chosen for the performance evaluation and analysis of the MTL-MGAN. The proposed algorithm has significantly improved the accuracy compared with related works. To examine the contributions of the individual components of the MTL-MGAN, ablation studies are conducted to confirm the effectiveness of the prioritization algorithm, the MTL, the negative transfer avoidance via loss functions, and the MGAN. The research implications are to confirm the feasibility of multiround transfer learning to enhance the optimal solution of the target model and to provide a generic approach to bridge the gap between the source domain and target domain using MGAN. [ABSTRACT FROM AUTHOR]

Virtual Private Networks (VPNs) are one example of encrypted communication services commonly used to bypass censorship and access geographically locked services. This study performed VPN and non-VPN traffic analysis and developed a classification system based on the new techniques of machine learning classifiers known as stacking ensemble learning. The methods used for VPN and Non-VPN classification use three machine learning techniques: random forest, neural network, and support vector machine. To assess the proposed method's performance, we tested it on a dataset containing 61 features. The experiment results accurately prove the study's classifiers to differentiate between VPN and Non-VPN traffic. The accuracy level was approximately 99% in the training and testing phase. The study's classifiers also show the best standard deviation, with a 100% accuracy rate compared to other A.I. classifier methods. [ABSTRACT FROM AUTHOR]

Smishing is a combined form of short message service (SMS) and phishing in which a malicious text message or SMS is sent to mobile users. This form of attack has come to be a severe cyber‐security difficulty and has triggered incredible monetary losses to the victims. Many antismishing solutions for mobile devices have been proposed till date but still, there is a lack of a full‐fledged solution. Therefore, this paper proposes an efficient approach that analyzes text content and uniform resource locator (URL) presented in the SMS. We have integrated the URL phishing classifier with the text classifier to improve accuracy as some of the SMS contain the URL with no text or much less text. To find out rare words in a report, depending upon the frequency of term (TF) and the reciprocal of document frequency TF‐inverse document frequency (IDF), a weighting framework TF‐IDF is used. We have used two data sets for both text as well as for URL phishing classifier and used a synthetic minority oversampling technique to balance the training data. The voting classifier simply merges the findings of each classifier passed into it and predicts the output on the basis of voting. In proposed approach integrating KNN, RF, and ETC can detect smishing messages with a 99.03% accuracy and 98.94% precision rate which is relatively efficient compared with existing ones like SmiDCA model which has the given accuracy of 96.40% using Random Forest classifier in BFSA, Feature‐Based it has an accuracy of 98.74% and 94.20% true positive rate and Smishing Detector it shows an overall accuracy of 96.29%. [ABSTRACT FROM AUTHOR]