Machine Learning-Based Attack Detection in Smart Home IoT Networks
Vinay Kumar Reddy Muthyala, Amit Lathigara, Nur Azaliah Abu Bakar · 2025
The Internet of Things, or IoT, is expanding quickly and improving our lives, but it also presents serious security risks. By providing an innovative anomaly detection system that uses ensemble learning to evaluate intrusion datasets and spot novel attack patterns, this work focuses on the security of smart home IoT devices. The study emphasizes how destructive IoT assaults are becoming more frequent and how anomalies could infect devices and lead to security lapses. The results show that, especially when used with the category smart home IoT dataset, the suggested machine learning model outperforms conventional techniques in identifying anomalies. By combining machine learning and statistical analysis, the study improves the security and dependability of smart home IoT systems. This method evaluates a number of factors and uses machine learning model scores to determine how trustworthy a gadget is. The method greatly improves the security of IoT devices, as demonstrated by real-world weather conditions and a publicly accessible smart home dataset.