LSTM (Long Short-Term Memory) based time sequence network anomaly detection method and device

Abstract

The invention provides LSTM (Long Short-Term Memory) based time sequence network anomaly detection method and device, and relates to the technical field of information security. The method comprises the following steps: acquiring an actual measured value of the to-be-detected network traffic; inputting the actual measured value of the to-be-detected network traffic into a LSTM based time sequencenetwork traffic prediction model, to obtain a predicted value of the to-be-detected network traffic; comparing the actual measured value of the to-be-detected network traffic with the predicted valueof the to-be-detected network traffic, to obtain an anomaly data detection result of the to-be-detected network traffic. The LSTM based time sequence network anomaly detection method provided by the invention can detect one-dimensional time sequence traffic data anomaly data and provide early warning in a large-scale network environment, thus improves the network anomaly detection efficiency, realizes a good effect of identifying network traffic anomaly, ensures relatively complete development fitting and can visually and obviously distinguish the anomaly information.

Description

technical field [0001] The present invention relates to the technical field of information security, in particular to an LSTM-based method and device for time series network anomaly detection. Background technique [0002] With the development of machine learning in recent years, there are many applications of machine learning algorithms in multidimensional outlier detection. One-dimensional time series outlier detection is still in its infancy, and exploration and research are relatively lacking. [0003] The manual method is not applicable to the detection of network outliers. The manual method can only find traffic information that is obviously different from the normal situation. It cannot judge the network anomalies that are not obvious, and the network traffic data volume is huge. unreasonable. [0004] For one-dimensional network outlier detection, the current common method is to use two methods: exploration based on the attributes of the data itself and time series...

AI Technical Summary

Problems solved by technology

In addition, LSTM-based systems can learn tasks such as translating languages, controlling robots, image analysis, document summarization, speech recognition image recognition, handwriting recognition, controlling chatbots, predicting diseases, click-through rates and stocks, synthesizing music, etc. The application of anomaly detection is in the early stage of exploration

Images