Key points are not available for this paper at this time.
Nowadays the security of computer devices is growing significantly. This is due to more and more devices areconnected to the network. For this reason, optimize the performance of systems able to detect intrusions (IDS) is a goalof common interest. The following work consists of use thegeneralizing power of neural networks to classify the attacks. In particular, we will use multilayer perceptron (MLP) withthe algorithm of back-propagation algorithm and the sigmoidalactivation function. We use a subset of the DARPA dataset, known as KDD99. It is a public dataset labeled for an IDS andpreviously processed. We will make an analysis of the resultsobtained using different configurations, varying the numberof hidden layers and the number of training epochs to obtaina low number of false results. We observe that it is requireda large number of training epochs and how, using the entiredata set consists of 41 features, the best classification is carriedout for the type of DOS and Probe attacks.
Amato et al. (Wed,) studied this question.