Classification And Accuracy Algorithm For Intrusion Detection Systems

1762 Words Oct 16th, 2016 8 Pages
The Network attribute selection, classification and accuracy algorithm (NASCA) is a three step algorithm for intrusion detection systems. The model first begins by extracting the information from the Transmission Control Protocol (TCP) and then ranking the relevant information that characterizes the network. During the second stage it classifies the network as to being under attack or not. Moreover, if the model detects that our system is under attack, it executes another classification approach to identify the type of the attack that is occurring. The third stage will provide us with results about the accuracy of the performed analytical estimate, or it will test the model how accurate is the reported guess of the network. The attribute selection method begins with choosing a subset of appropriate information by removing redundant, unrelated and noisy data from the original dataset. In real-world, the representation of data often includes an abundance of features, however, only a limited number of them, may be associated with the essential idea. The classification method starts initially with an existing set of labeled networks, consequently it learns the dependency between the content of the network and its corresponding label and then predicts the label of a set of unlabeled networks as most accurate as possible using a decision tree type of analysis.
The common classification procedures are based on familiar machine learning models as Naive Bayes or discriminant models…

Related Documents