Natural Images Allow Universal Adversarial Attacks on Medical Image Classification Using Deep Neural Networks with Transfer Learning. [PDF]
Minagi A, Hirano H, Takemoto K.
europepmc +1 more source
MLDAS: Machine Learning Dynamic Algorithm Selection for Software‐Defined Networking Security
ABSTRACT Network security is a critical concern in the digital landscape of today, with users demanding secure browsing experiences and protection of their personal data. This study explores the dynamic integration of Machine Learning (ML) algorithms with Software‐Defined Networking (SDN) controllers to enhance network security through adaptive ...
Pablo Benlloch +3 more
wiley +1 more source
A multi-layered defense against adversarial attacks in brain tumor classification using ensemble adversarial training and feature squeezing. [PDF]
Yinusa A, Faezipour M.
europepmc +1 more source
An AI-driven solution to prevent adversarial attacks on mobile Vehicle-to-Microgrid services
Ahmed M. Omara, Burak Kantarcı
openalex +1 more source
HADA: A Hybrid Anomaly Detection Approach Using Unsupervised Machine Learning
Overview of HADA, an unsupervised fraud detection pipeline that preprocesses and scales transaction data, applies PCA for dimensionality reduction, scores anomalies using Isolation Forest, selects anomalous transactions via thresholding, and clusters the selected anomalies using Agglomerative Hierarchical Clustering (AHC) to produce interpretable ...
Francis Thiong'o +3 more
wiley +1 more source
Channel Attention for Fire and Smoke Detection: Impact of Augmentation, Color Spaces, and Adversarial Attacks. [PDF]
Ejaz U, Hamza MA, Kim HC.
europepmc +1 more source
A novel three‐layer architecture is proposed for cell‐to‐cell communication, integrating molecular, nano‐electronic, and digital components. It enables bio‐integrated applications such as remote emotional touching, consensual telepathy, and autonomous drug delivery. Molecular Communication (MC) is utilized with advanced implants, modulation techniques,
Saeed Banaeian Far +2 more
wiley +1 more source
Salient object detection dataset with adversarial attacks for genetic programming and neural networks. [PDF]
Olague M +3 more
europepmc +1 more source
Overview of the proposed work. ABSTRACT Identifying cyber threats maintains the security and operational stability of smart grid systems because they experience escalating attacks that endanger both operating data reliability and system stability and electricity grid performance.
Priya R. Karpaga +3 more
wiley +1 more source

