A hybrid BGWO with KPCA for intrusion detection
Velliangiri Sarveshwaran · Journal of Experimental & Theoretical Artificial Intelligence · 2019
Intrusion detection is the primary model for giving security to the network. There are numerous issues with conventional intrusion detection models (for example, low detection ability against obscure arranges attack, high false caution rate, and inadequate investigation ability). The real challenging in design IDS with enhanced precision and diminished the training time. This paper come up with hybrid intrusion detection model by incorporating the kernel principal component analysis (KPCA) and binary grey wolf optimization (BGWO) with support vector machine (SVM). The curiosity of the paper is the headway of part parameters of the SVM classifier using the customized parameter assurance procedure. This proposed hybrid method streamlines the discipline factor (C) and kernel parameters (σ) and the size of tube ε of SVM, like this enhancing the precision of the classifier and reduces the testing and training time.