Data anomaly detection method and device based on similar slow feature transformation vector

Abstract

The invention discloses a data anomaly detection method and device based on similar slow feature transformation vectors, and the method comprises the steps: carrying out the standardization processing of to-be-detected real-time data, so as to obtain standardized data; statistical magnitude is generated based on the standardized data, a detection sample is constructed by using the statistical magnitude, and the statistical magnitude is a similar slow feature transformation vector; taking the detection sample as input of a support vector data description model, and determining a first distance between the detection sample and the sphere center of the hypersphere determined based on the support vector data description model; and determining that the real-time data to be detected is abnormal according to the fact that the first distance is greater than the second distance. Real-time data are described based on similar slow feature transformation vectors, whether the data are abnormal or not can be judged according to the distance between the detection sample determined by the support vector data description model and the center of the hypersphere, and therefore whether large-scale system data and dynamic data are abnormal or not can be effectively detected.

Description

technical field [0001] The present invention relates to the technical field of cloud computing monitoring, and more specifically, the present invention can provide a data anomaly detection method and device based on similar slow feature transformation vectors. Background technique [0002] With the continuous development of information technology, the scale of various software systems and hardware systems is increasing day by day, so it is necessary to effectively monitor them, so as to deal with them in time in case of abnormalities and ensure the stability of the operating status of each system. With the advancement of database technology, various system data can be stored, so multivariate statistical data detection techniques have been widely used, such as principal component analysis (PCA) and so on. However, due to the continuous expansion of the system scale and the increasingly complex connections between the various components of the system, the traditional detection...

AI Technical Summary

Problems solved by technology

[0003] In order to solve the problem that the traditional solution cannot effectively detect whether the data is abnormal, the present invention provides a data anomaly detection method and device based on similar slow feature transformation vectors, so as to achieve the technical purpose of effectively detecting whether the data is abnormal

Images