首页|Anderson University Researchers Report on Findings in Machine Learning (TXAI-ADV : Trustworthy XAI for Defending AI Models against Adversarial Attacks in Realist ic CIoT)
Anderson University Researchers Report on Findings in Machine Learning (TXAI-ADV : Trustworthy XAI for Defending AI Models against Adversarial Attacks in Realist ic CIoT)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators discuss new findings in artificial intelligence. According to news reporting from Anderson, South Caroli na, by NewsRx journalists, research stated, "Adversarial attacks are more preval ent in Consumer Internet of Things (CIoT) devices (i.e., smart home devices, cam eras, actuators, sensors, and micro-controllers) because of their growing integr ation into daily activities, which brings attention to their possible shortcomin gs and usefulness." The news correspondents obtained a quote from the research from Anderson Univers ity: "Keeping protection in the CIoT and countering emerging risks require const ant updates and monitoring of these devices. Machine learning (ML), in combinati on with Explainable Artificial Intelligence (XAI), has become an essential compo nent of the CIoT ecosystem due to its rapid advancement and impressive results a cross several application domains for attack detection, prevention, mitigation, and providing explanations of such decisions. These attacks exploit and steal se nsitive data, disrupt the devices' functionality, or gain unauthorized access to connected networks. This research generates a novel dataset by injecting advers arial attacks into the CICIoT2023 dataset. It presents an adversarial attack det ection approach named TXAI-ADV that utilizes deep learning (Mutli-Layer Perceptr on (MLP) and Deep Neural Network (DNN)) and machine learning classifiers (K-Near est Neighbor (KNN), Support Vector Classifier (SVC), Gaussian Naive Bayes (GNB), ensemble voting, and Meta Classifier) to detect attacks and avert such situatio ns rapidly in a CIoT. This study utilized Shapley Additive Explanations (SHAP) t echniques, an XAI technique, to analyze the average impact of each class feature on the proposed models and select optimal features for the adversarial attacks dataset."
Anderson UniversityAndersonSouth Car olinaUnited StatesNorth and Central AmericaCyborgsEmerging TechnologiesMachine Learning
NETL
NSTL
