Abnormity detection method for enterprise industry classification

Abstract

The invention discloses an abnormity detection method for enterprise industry classification, which comprises the following steps of: firstly, extracting to-be-mined text and non-text information in taxpayer industry information, and carrying out feature processing and coding processing; Secondly, constructing a deep network structure conforming to the industry classification abnormity detection problem, and determining the number of neurons of an input layer and an output layer of the network according to the characteristic dimension of the coded data; Thirdly, on the basis of the constructeddeep network structure, adopting different training strategies to train the industry large-class network and the industry detail network through cross validation; And finally, carrying out abnormitydetection on the industry large class by using dimension reduction characteristics of the industry large class network in combination with an SOS abnormity detection algorithm, and carrying out abnormity detection on industry details according to reconstruction characteristics of the industry detail network. According to the invention, the TADM model is utilized to carry out abnormal detection onthe original data, and macroscopic management work such as national statistics, tax collection and industrial and commercial management can be analyzed more reasonably and accurately.

Description

technical field [0001] The invention belongs to the field of data mining, in particular to a TADM (Two-level Anomaly Detection Model, 2-level anomaly detection model)-oriented anomaly detection method for industry classification of enterprises. Background technique [0002] After the reform and opening up, my country's national economy has developed rapidly, the market economy has continued to prosper, the country's economic structure has gradually improved, and the division of labor among enterprises has gradually become more refined. In the new era, the study of enterprise industry classification plays a fundamental role in promoting finance, taxation, and national standard management, and also provides a basis for further analysis of national economic industries and industrial development status, and grasping national economic development trends. The "National Economic Industry Classification" (GB / T 4754-2017) issued by the General Administration of Quality Supervision, I...

AI Technical Summary

Problems solved by technology

[0010] Although the above traditional methods can solve their specific anomaly detection problems, it is difficult to directly extend to the anomaly detection problem of industry classification, because the anomaly detection of industry classification has the characteristics of multi-category and multi-level

First, enterprise industry classification is a multi-classification problem, and the variety of categories and the amount of data make the problem of anomaly detection complicated.

The self-encoder network structures in Documents 1 and 3 are too simple, with only one hidden layer, unable to effectively extract detailed features of the data, and severely lack generalization ability under large-scale data sets; Document 2 defines local outliers by using k-neighborhood dispersion coefficient, but in the case of many industry categories and a large amount of industry information data, the selection of k value becomes extremely difficult

Second, there is a hierarchical affiliation relationship between the industry category and the details of the enterprise. The two belong to different levels, and the industry details are the expansion and refinement of the industry category. Any enterprise corresponds to an industry category and a Industry details, each requires data with different information granularity (reflecting the level of detail of the information) for anomaly analysis. Documents 1-3 do not have a solution for the multi-level anomaly detection problem of industry classification

Images