Publication

Abstract : An Intrusion Detection System may keep an eye on a network and its systems to look for any indications of an attack or a violation of security protocols. It’s a measure you can take to prevent malicious actors from accessing your system. You may improve your ability to identify threats and reduce the number of times they sneak past your defenses by combining binary classification with the k-nn approach. In a further step, we use binary classifiers and a single aggregation module to establish the appropriate labels for each node in the network. Once the proper classes have been identified, the links containing confusing classes are passed on to the next stage, when k-nn techniques are employed to further untangle the classes. In order to identify and categories abnormal connections, a multiclass support vector machine method is utilised. When applied to the NSL-KDD dataset, the proposed method yields impressive results.