A Hybrid Machine Learning Model for Network Intrusion Detection Focusing on NSL-KDD Dataset
The Internet of Things and numerous social networking sites accumulate massive network traffic along with a lot of content due to the massive rise of technology including such cloud technology. There is always the prospect of threats through intruders due to the extreme volume of that same server and records, which may be due to severe harmful behaviors. So that the whole network operation should also be supervised as well as intruders intercepted early. The proposed WRFSVM framework incorporates that fusion scheme utilizing dual classifiers firstly A hybrid classifier based on weighted random forest including A Decision Tree-based attribute Weighted Average One-Dependence Estimator classifier and secondly the support vector approach. This proposed WRFSVM includes the "Recursive Feature Omission" approach through the reliable component selection and DTAODE that resolve problems connected to attribute dependence. Hybrid Random Forest, as well as the SVM method, extends after the extraction of features from those in the set of data. Another greatest strength of just the proposed model has always been its comprehensive range among attributes and hence the methodology towards integrated classification leads to increased precision, recall as well as accuracy. WEKA simulator platforms, as well as the NSL-KDD set of data, were used in the sample through experimental exploration throughout this research.