Fault early warning algorithm based on outlier mining

Abstract

The invention discloses a fault early warning algorithm based on outlier mining, which comprises the following steps of S1.carrying out attribute outlier detection and judging whether data points are outliers one by one according to the attributes of a data object; S2.detecting and marking abnormal attributes of the data object according to the definition of the abnormal attributes; S3.seperating and outputting the data of the outliers according to abnormal marks of the data object; and S4.carrying out timing outlier detection, combining a plurality of adjacent sequence points, detecting through the abnormal mode structure or the abnormal frequency and making characteristics of the abnormal data remarkable to complete fault early warning. The fault early warning algorithm fully considers the importance of the anomaly of the single attribute on judgment of the fault data for the abnormal detection of the attributes, the method for timing outlier detection fully considers the natural change trend of the data of a website with the change of time, the fault early warning algorithm is very suitable for the fault early warning of a rapid development stage of the business, and the method is simple and high in efficiency.

Description

technical field [0001] The invention relates to the technical field of fault early warning algorithms, in particular to a fault early warning algorithm based on heterogeneous mining. Background technique [0002] With the advancement of science and the development of society, more and more modern equipment has entered people's lives. Modern equipment not only makes people's lives more convenient, but also promotes social progress. [0003] The failure operation mode of the website is an atypical operation mode in the operation of the website, so the database of the website operation data may contain some data objects, which are inconsistent with the general behavior or model of the data, and these data objects are abnormal points or called heterogeneous points . [0004] Under the prior art, many data mining methods minimize the influence of such data points, or exclude such data points as noise. However, for the fault warning system, heterogeneous points are very importan...

AI Technical Summary

Problems solved by technology

However, for the fault warning system, heterogeneous points are very important, and important information faults are hidden in them, which indicates the operation failure of the system. If these heterogeneous points are eliminated, it is not conducive to the troubleshooting of website faults, and even causes the website to be paralyzed.