Modified Intrusion Detection Tree with Hybrid Deep Learning Framework based Cyber Security Intrusion Detection Model

In modern era, the most pressing issue facing society is protection against cyberattacks on networks. The frequency of cyber-attacks in present world makes problem providing feasible security to computer system from potential risks important and crucial. Network cannot be effectively monitored protected without use intrusion detection systems (IDSs). DLTs (Deep learning methods) MLTs (machine techniques) are being employed information domains for building IDSs. These IDSs capable automatically timely identifying harmful attacks. IntruDTree (Intrusion Detection Tree), a model based that detects attacks effectively, shown existing research effort. This model, however, suffers an overfitting problem, which occurs when method perfectly matches training data but fails generalize new data. To address issue, this study introduces MIntruDTree-HDL (Modified with Hybrid Deep Learning) framework, improves performance prediction framework predicts classifies cyber assaults network using M-IntruDtree IDS Tree) CRNNs (convolution recurrent neural networks). rank key characteristics, first create modified tree-based generalized M-IntruDTree. CNNs networks) then convolution collect local information, while RNNs (recurrent capture temporal features increase prediction. not only accurate predicting unknown test scenarios, it also results reduced computational costs due its dimensionality reductions. efficacy suggested schemes was benchmarked cybersecurity datasets terms precisions, recalls, fscores, accuracies, ROC. simulation show proposed outperforms current approaches, high rate malicious attack accuracy.