Statistical methods based on local and global feature differences in a single individual

Abstract

The invention provides a statistical method based on the difference between local and whole features in a single individual, which comprises the following steps: local feature statistics is used to select and recombine the feature set of the local working state in the original data to obtain the local feature statistical result; the local feature statistics is used to obtain the local feature statistical result. The whole feature statistic is obtained by choosing the feature set which does not include part of the whole working state in the original data. Difference statistics, through the difference function to calculate the difference between the local feature statistics results and the overall feature statistics results, to get the difference results; Statistical explicitness, through statistical methods for statistical analysis of the difference between the results. The invention reorganizes and analyzes the single statistical result representing the key characteristic of the localworking state and the statistical result representing the key characteristic of the whole working state at a certain time, calculates the difference of the working state between the local and the whole, and thus obtains the judging basis of whether the local working state of the individual is normal or not.

Description

technical field [0001] The invention relates to the field of computer science and technology, and in particular to a statistical method, electronic equipment, and a storage medium based on local and overall feature differences in a single individual. Background technique [0002] All partial components work in the system and are in the same working environment. In order to judge whether the working state of a part is normal at a certain moment, it is not perfect to only use a single local data feature and the corresponding characteristic index to compare and judge the working state. At the same time, if the characteristic indicators of the working status of other parts (or even the whole) are very different, then it is very likely that the part has not worked properly or even failed. Therefore, it is necessary to count the characteristic indicators representing the working status of the part and the whole under the same system at the same time, and analyze the differences b...

AI Technical Summary

Problems solved by technology

[0003] In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide a statistical method based on the difference of local and overall features in a single individual, which solves the problem of comparing and judging only local single data features and corresponding feature indicators in the prior art The working state method is not perfect for the problem

Images