A Low-Rank Learning-Based Multi-Label Security Solution for Industry 5.0 Consumers Using Machine Learning Classifiers

Abstract

The need for networking in smart industries known as Industry 5.0 has grown critical, and it is especially important for the security and privacy of the applications. To counter threats to important consumers devices’ sensitive data, various applications of smart industries require intelligent schemes and architectures. The data which is recorded and stored is vulnerable to security breaches. These attacks, though, can be recognized using machine-learning approaches, which necessitate the construction of a new dataset. The following paper uses a hybrid intrusion dataset which is used to solve multi-label classification problems using a multi-criteria decision-making process for consumer devices. The use of two datasets having the same attack from two different classes is difficult to recognize the class of attack. Our proposed model is going to recognize the type of attack from the two classes by combining the machine learning classifiers with the multiple criteria decision-making process and validated over existing state of art techniques.