Network anomaly detection using artificial neural networks
Sergey Andropov, Alexei Valerievich Guirik, Mikhail Yurievich Budko, Marina Borisovna Budko · 2017
This paper presents a method of identifying and classifying network anomalies using an artificial neural network for analyzing data gathered via Netflow protocol. Potential anomalies and their properties are described. We propose using a multilayer perceptron, trained with the backpropagation algorithm. We experiment both with datasets acquired from a real ISP monitoring system and with datasets modified to simulate the presence of anomalies; some Netflow records are modified to contain known patterns of several network attacks. We evaluate the viability of the approach by practical experimentation with various anomalies and iteration sizes.