Genetic Algorithm-Conditional Mutual Information Maximization based feature selection for Bot Attack Classification in IoT devices

Abstract

The evolution of computing is increasing in a vast manner that will integrate many physical objects and the internet to generate a new interconnection, such as the Internet of Things (IoT). It is estimated that the number of devices that will be interconnected to the internet will be more than trillions until 2025. Due to the lack of interoperability when these devices are interconnected in a vast heterogeneous network, it is tough to define and apply security mechanisms. The IoT networks have been exposed to many vulnerable attacks that disturb the network. Therefore, designing an intrusion detection system that provides additional security tools specific to IoT is needed to apply security mechanisms to detect the attacks in the network. In this paper, we propose a novel hybrid GA-CMIM machine learning algorithm that improves the efficiency in detecting the botnet intrusions with the set of optimal features that are selected from the dataset using a feature selection method.

An Improved Mutual Information Feature Selection Technique for Intrusion Detection Systems in the Internet of Medical Things

“…This method measures the dependence between each feature and the class label and selects the features with the highest mutual information scores [ 44 ]. The main advantage of using mutual information is that it considers the nonlinear relationships between features and class labels, making it suitable for handling complex and nonlinear data patterns in IoT environments [ 48 , 49 ]. Furthermore, mutual information is also able to handle noisy and incomplete data, which are commonly encountered in IoT networks [ 48 , 50 ].…”

“…By selecting the most relevant and informative features, mutual information feature selection can effectively reduce the dimensionality of the data and improve the accuracy of the IDS. Several studies have demonstrated the effectiveness of mutual information feature selection for IDS in IoT environments and have compared its performance with other feature selection methods [ 45 , 49 , 51 ].…”