Anomaly detection method and system

Abstract

The embodiment of the invention provides an anomaly detection method and system, and the method comprises the steps: obtaining time sequence data of each index in a preset period before the current time on a to-be-detected platform, and carrying out the preprocessing, and obtaining a real value time sequence of the index; for each index, using no less than two machine learning models to predict the real value time sequence of the index to obtain a predicted value time sequence corresponding to each machine learning model of the index in the preset period, and using the predicted value time sequence closest to the real value time sequence as the predicted value time sequence of the index; comparing the predicted value time sequence with the true value time sequence of each index to obtain an error value time sequence; judging each error value in a specified time period according to the error value time sequence, and judging whether a time point corresponding to each error value is an abnormal point or not; and giving an alarm for abnormal points. The predicted value time sequence is calculated through the machine learning model, the calculation amount is reduced, and the calculation time is short.

Description

technical field [0001] The invention relates to the field of intelligent detection, in particular to an abnormality detection method and system. Background technique [0002] At present, the open source anomaly detection system based on time series mainly uses ARIMA, but ARIMA detects time series based on the trained model. Millions of time series are detected, and the longer the period, the slower the detection speed, and the results cannot be calculated even for several hours. Contents of the invention [0003] The embodiment of the present invention provides a method and system for anomaly detection, which does not need to fit a complex model, and only needs to use a machine learning model to calculate the time series of predicted values, which greatly reduces the amount of calculation and can effectively reduce the calculation time of the prediction model . [0004] To achieve the above purpose, on the one hand, an embodiment of the present invention provides an anom...

AI Technical Summary

Problems solved by technology

[0002] At present, the open source anomaly detection system based on time series mainly uses ARIMA, but ARIMA detects time series based on the trained model. The detection of millions of time series, and the longer the period, the slower the detection speed, and the results cannot be calculated even for several hours

Images