Review on machine learning and deep learning algorithms for IoT security

With its rapid expansion in more sectors, for instance, wearables, smart sensors, plus house devices, the Internet of Things (IoT) is drifted to have a significant influence on many parts of our life. IoT devices stand out for their connectivity, ubiquity, and low processing power. By 2025, there will likely be 30.9 billion devices adjoined to the Internet, since the count of IoT devices in use worldwide is growing quickly. This eruption about IoT devices, which in analogy to desktop PCs, can be quickly increased, has caused an increase in occurrences of IoT-based cyber intrusions. It is necessary to create new methods for identifying attacks launched from hacked IoT devices in order to address this challenge. The best detective control solution against attacks caused by IoT devices, in this context, uses machine and deep learning approaches. This paper attempts some analysis of technologies, threats arising from IoT devices, and intrusion detection system overview as they associate with IoT systems. The investigation of several machine learning plus deep learning concepts appropriate for identifying IoT devices linked with cyberattacks is also included in this paper.