Providing a method to reduce the false alarm rate in network intrusion detection systems using the multilayer perceptron technique and backpropagation algorithm

Abstract

Along with the growth of computer networks, attacks have spread to these networks and are carried out in many ways. Influencing the collection of illegal acts that compromise the integrity or access to a source. Normal behaviors use their own patterns in network traffic data, which can be exploited by using different data mining techniques. Traditional methods cannot effectively exploit the discovery of unknown patterns, as human resources encounter a network of computer systems that have high speed and complexity while performing intrusion detection analysis. In this paper, a new method for intrusion detection using multi-layer perceptron (mlp) neural network is presented. The proposed method consists of two phases, the first phase of training the classification parameters using training data, and the second phase of classifying the classification testing data. In mlp neural network, backpropagation error Algorithm are used for training. The use of mlp is effective in decreasing the false positive rate. The results in the proposed method are much better than other methods.