Neural Networks in Predictive Analytics for the IIoT Systems
Maryna Kolisnyk, Axel Jantsch · Advances in computational intelligence and robotics book series · 2025
The Industrial Internet of Things (IIoT) is a complex system that consists of many components, each of which is exposed to the risk of fault and failure of both hardware and software (SW) due to various reasons (cyber-attacks on vulnerabilities, different types of defects, etc.). The application of predictive analytics based on Artificial Intelligence (AI) techniques in particular neural networks to complex systems such as IIoT will not only allow us to quickly and accurately identify the nature and location of faults, but also to predict them in the future by applying classification and regression techniques. In this section, a detailed analysis, systematisation and categorisation of AI methods: machine learning and deep learning methods, neural networks and their activation functions in implementing AI-based predictive analytics (PA) for IIoT systems and their components has been carried out. This study will help in the development of PA methods for IIoT systems, researchers in the field of and manufacturers of SW for PA realisation.