The Internet of Things (IoT) is revolutionizing industries by enabling seamless interconnectivity across domains such as healthcare, smart cities, the Industrial Internet of Things (IIoT), and the Internet of Vehicles (IoV). However, IoT security remains a significant challenge due to vulnerabilities related to data breaches, privacy concerns, cyber threats, and trust management issues. Addressing these risks requires advanced security mechanisms, with machine learning (ML) emerging as a powerful tool for anomaly detection, intrusion detection, and threat mitigation. This survey provides a comprehensive review of ML-driven IoT security solutions from 2020 to 2024, examining the effectiveness of supervised, unsupervised, and reinforcement learning approaches, as well as advanced techniques such as deep learning (DL), ensemble learning (EL), federated learning (FL), and transfer learning (TL). A systematic classification of ML techniques is presented based on their IoT security applications, along with a taxonomy of security threats and a critical evaluation of existing solutions in terms of scalability, computational efficiency, and privacy preservation. Additionally, this study identifies key limitations of current ML approaches, including high computational costs, adversarial vulnerabilities, and interpretability challenges, while outlining future research opportunities such as privacy-preserving ML, explainable AI, and edge-based security frameworks. By synthesizing insights from recent advancements, this paper provides a structured framework for developing robust, intelligent, and adaptive IoT security solutions. The findings aim to guide researchers and practitioners in designing next-generation cybersecurity models capable of effectively countering emerging threats in IoT ecosystems.

The Internet of Medical Things (IoMT) is revolutionizing healthcare by enabling advanced patient care through interconnected medical devices and systems. However, its critical role and sensitive data make it a prime target for cyber threats, requiring the implementation of effective security solutions. This paper presents a novel intrusion detection system (IDS) specifically designed for IoMT networks. The proposed IDS leverages machine learning (ML) and deep learning (DL) techniques, employing a stacking ensemble method to enhance detection accuracy by integrating the strengths of multiple classifiers. To ensure real-time performance, the IDS is implemented within a Kappa Architecture framework, enabling continuous processing of IoMT data streams. The system effectively detects and classifies a wide range of cyberattacks, including ARP spoofing, DoS, Smurf, and Port Scan, achieving an outstanding detection accuracy of 0.991 in binary classification and 0.993 in multi-class classification. This research highlights the potential of combining advanced ML and DL methods with ensemble learning to address the unique cybersecurity challenges of IoMT systems, providing a reliable and scalable solution for safeguarding healthcare services.

Background: Breast cancer screening through mammography interpretation is crucial for early detection and improved patient outcomes. However, the manual classification of mammograms using the BIRADS (Breast Imaging-Reporting and Data System) remains challenging due to subtle imaging features, inter-reader variability, and increasing radiologist workload. Traditional computer-aided detection systems often struggle with complex feature extraction and contextual understanding of mammographic abnormalities. To address these limitations, this study proposes MammoViT, a novel hybrid deep learning framework that leverages both ResNet50’s hierarchical feature extraction capabilities and Vision Transformer’s ability to capture long-range dependencies in images. Methods: We implemented a multi-stage approach utilizing a pre-trained ResNet50 model for initial feature extraction from mammogram images. To address the significant class imbalance in our four-class BIRADS dataset, we applied SMOTE (Synthetic Minority Over-sampling Technique) to generate synthetic samples for minority classes. The extracted feature arrays were transformed into non-overlapping patches with positional encodings for Vision Transformer processing. The Vision Transformer employs multi-head self-attention mechanisms to capture both local and global relationships between image patches, with each attention head learning different aspects of spatial dependencies. The model was optimized using Keras Tuner and trained using 5-fold cross-validation with early stopping to prevent overfitting. Results: MammoViT achieved 97.4% accuracy in classifying mammogram images across different BIRADS categories. The model’s effectiveness was validated through comprehensive evaluation metrics, including a classification report, confusion matrix, probability distribution, and comparison with existing studies. Conclusions: MammoViT effectively combines ResNet50 and Vision Transformer architectures while addressing the challenge of imbalanced medical imaging datasets. The high accuracy and robust performance demonstrate its potential as a reliable tool for supporting clinical decision-making in breast cancer screening.

The Controller Area Network (CAN) bus has been implemented in most modern Vehicles. Various attacks can be launched against the CAN bus protocol because it is designed without security mechanisms. It is essential to develop a highly accurate intrusion detection system (IDS) for CAN bus attacks. We design an effective ensemble learning-based IDS scheme for detecting and classifying DoS, fuzzing, replay, and spoofing attacks. These are common CAN bus attacks that can threaten the safety of a vehicle’s driver, passengers, and pedestrians. For this purpose, we utilize supervised machine learning in combination with ensemble methods. We first perform data balancing and feature selection. We build and fine-tune random forest, Xtreme gradient boosting, and decision tree supervised learning models. We then combine these models with voting, stacking, and bagging ensemble methods. The results obtained demonstrate the effectiveness of the proposed scheme when trained on real-life CAN traffic datasets to detect and classify these four attacks. The stacking method achieved the highest performance in terms of accuracy, precision, recall, F1-score, and area-under-the-curve receiver operator characteristic (ROC-AUC). The stacking method outperformed other recently proposed methods with an F1-score, precision, recall, and accuracy of 0.993, 0.993, 0.993, and 0.986, respectively.