Разработка модели обнаружения сетевых атак на основе искусственной нейронной сети
The article describes the development of a neural network for the detection and classification of network attacks. The paper estimated the importance of the input parameters of compounds, represents built a neural network with a full and reduced set of parameters, depicts a completed comparative analysis of their performance. The result is the optimal model of the neural network, which has the 21 input parameter and can determine the type of attack with a probability of 99.83%. Comparison of the resulting neural network model with the same neural networks shows high efficiency of designed neural network. The resulting neural network can be used successfully as a component of the intrusion detection system.