Deep learning aplications for network intrusion detection

Network intrusion detection systems are a critical part of today's information systems, it has started to rise as an essential research area, and the use of artificial neural networks has become increasingly popular in this area. However, there are deficiencies in the comparison studies about artificial neural network architectures and their parameters. In this study, it is aimed to create a basis for future research and academic studies with artificial neural network architectures used in the field of network intrusion detection systems. In this respect, KDD-99, which is accepted as the benchmark data set, was used. Feedforward artificial neural network architecture and convolutional neural network architecture have been used on this dataset, and over 90% success has been achieved, and the results have been compared.