Analysis of Intrusion Detection Systems in UNSW-NB15 and NSL-KDD Datasets with Machine Learning Algorithms

Recently, the need for Network-based systems and smart devices has been increasing rapidly. The use of in almost every field, provision services by private public institutions over network servers, cloud technologies database are completely remotely controlled. Due to these requirements systems, malicious software users, unfortunately, their interest areas. Some organizations exposed hundreds or even thousands attacks daily. Therefore, it is not enough solve with a virus program firewall. Detection correct analysis vital operation entire system. With deep learning machine learning, attack detection classification can be done successfully. In this study, comprehensive process was performed on UNSW-NB15 NSL-KDD datasets existing algorithms. UNSW-NB115 dataset, 98.6% 98.3% accuracy were obtained two-class multi-class, respectively, 97.8% 93.4% dataset. results prove that algorithms lateral solution intrusion systems.