Optimal DBN‐based distributed attack detection model for Internet of Things

Abstract

SummaryThis paper introduces a new detection mechanism for defending the cyberspace with a new logic that aiding the concept of deep learning. The process involves two phases, namely, feature extraction and classification. The initial phase is the feature extraction, in which the features are extracted from the given input data by the renowned principal component analysis (PCA). Subsequently, the extracted features are subjected to the classification phase, where the deep belief network (DBN) model is used. The DBN model classifies the presence of attacks like denial of service (DoS), probe, R2L, and U2R. In order to make the performance more excellent, this paper diverts the strategy to a new concept termed “Optimization Concept.” Here, the hidden neuron of DBN is optimally selected by a new algorithm termed novel mutation rate‐based lion algorithm (NMR‐LA), which is the modified model of lion algorithm (LA). The performance of proposed algorithm NMR‐LA is compared over the conventional models in terms of both positive and negative measures like accuracy, sensitivity, specificity, precision, negative predictive value (NPV), F1 score and Mathews correlation coefficient (MCC), false‐positive rate (FPR), false‐negative rate (FNR), and false‐discovery rate (FDR) and proves the betterments of proposed work.