Machine Learning for IoT Device Security in Smart Homes

作者：Mohammad Abu Kausar, Yousif K. Yousif, M. Batahari, M. Nasar, Nidal A. Al-Dmour, Ali Q. Saeed · 发表于：2025 3rd International Conference on Business Analytics for Technology and Security (ICBATS) · 年份：2025 · DOI：10.1109/icbats66542.2025.11258288 · 被引用次数：3

The massive rise of smart home Internet of Things (IoT) devices has provided immense comfort and automation, but it has also led to greater security vulnerabilities. Because of these limitations, the multidimensional protocols as well as decentralized nature of these devices, ensuring the security of such devices becomes an unavoidable issue. In this article, we are discussing how the application of machine learning (ML) techniques can help in improving IoT device security in smart homes. The study employs algorithms of ML in real time to detect and mitigate possible faults such as unauthorized access, dysfunctional behavior, or data leaks. A comprehensive pattern of IoT ecosystems combining supervised, unsupervised, and reinforcement learning methods is proposed to be the practical solution. Finally, we have conducted experiments using both synthetic and real-world IoT datasets which have shown that such ML models can be successfully utilized for various applications including but not limited to intrusion detection, malware classification, and anomaly detection with high accuracy rates. The obtained findings verify that and show us how to utilize the intelligence of ML to interactively protect IoT devices at the same time ensuring their productivity, security, and privacy. Machine learning plays the crucial part of the project called “Next Generation Smart Homes: Resilience and Security”.