Identification method and device for abnormal data of business expansion project, equipment and storage medium

Abstract

The invention relates to a method for identifying abnormal data of a business expansion project, and the method comprises the following steps: obtaining business expansion project data which comprises the number of project process nodes, the working time of the nodes, and the project cost; a t-distributed random neighborhood embedding algorithm t-SNE is adopted to carry out feature recombination and dimension reduction on the business expansion project data; constructing an isolated forest model; and inputting the dimension-reduced business expansion project data into the isolated forest model to perform anomaly detection on the business expansion project data to obtain a detection result. The method focuses on research on anomaly detection under unsupervised learning, dimension reduction and recombination are carried out on high-dimension data through the t-SNE algorithm, and then abnormal data are detected by using the isolated forest. The isolated forest is an unsupervised rapid anomaly detection method based on combination, has linear time complexity and high precision, can achieve a good anomaly detection effect only by a small number of samples, and has the advantage of rapid convergence.

Description

technical field [0001] The present invention relates to the technical field of abnormal business expansion item data, in particular to a SOM-E-based business expansion project clustering method, device, equipment and storage medium. Background technique [0002] With the rapid development of my country's power grid industry, how to improve power quality and provide high-quality power services has always been a concern in the power grid field. [0003] In the industry expansion report and installation process, the problem that often occurs is that the data source of the industry expansion supporting project often causes a small amount of data to be abnormal due to various reasons. Expand the abnormal data identification of supporting project data. However, most of the current anomaly identification methods are to identify the imbalance of the data, and because the high-dimensional data will be more sparse in the spatial distribution than the low-dimensional data, so it is mo...

AI Technical Summary

Problems solved by technology

However, most of the common anomaly identification methods now are to identify the imbalance of the data, and because the high-dimensional data is more sparse in spatial distribution than the low-dimensional data, it is more difficult to identify, thus ignoring the high-dimensional data. Factors Affecting Outlier Identification

In turn, the identification of abnormal data is not accurate enough and the sensitivity is not high

Images