An intrusion detection approach based on incremental long short-term memory

The notorious attacks of the last few years have propelled cyber security to the top of the boardroom agenda, and raised the level of criticality to new heights. Therefore, building a secure system has become an important issue that cannot be delayed. In this paper, we propose an intrusion detection approach based on incremental long short-term memory to detect attacks. In order to capture the dynamic information of traffic, we introduce increment which is calculated as the product of function and derivative to long short-term memory (LSTM). Furthermore, the state change are applied to LSTM which is considered as incremental LSTM. Finally, we analyzed the effect of the state change on the performance of incremental LSTM by experiments. Experiments show that the intrusion detection method based on incremental LSTM has a higher accuracy than other methods.