Abnormal case identification method and device based on tree structure, equipment and storage medium

Abstract

The invention discloses an abnormal case identification method and device based on a tree structure, equipment and a storage medium, and belongs to the technical field of artificial intelligence, andthe method comprises the steps: obtaining an original training sample from an initial case database; calculating an abnormal score of the original training sample based on an IForest algorithm; comparing the abnormal score of the original training sample with a preset threshold, classifying the original training sample according to a comparison result, and forming a target training set; performingmodel training on the initial recognition model through the target training set, and outputting an abnormal recognition model; and importing the case data of the to-be-identified case into the exception identification model, and outputting an identification result. The invention also relates to a blockchain technology. The abnormal score of the original training sample can be stored in the blockchain. The abnormal score of the original training sample is calculated based on the IForest algorithm, and the original training sample is classified according to the abnormal score so that the interference of human factors is effectively eliminated, and the accuracy of the case abnormality recognition model is improved.

Description

technical field [0001] The application belongs to the technical field of artificial intelligence, and specifically relates to a method, device, equipment and storage medium for identifying abnormal cases based on a tree structure. Background technique [0002] Traditional identification of case abnormalities often adopts the methods of surveyors and statistical models, but both of them use human judgment as the benchmark, and human judgment is difficult to quantify and has subjective factors. And in the process of human judgment, abnormal cases are usually marked as positive samples, and other cases are considered as non-abnormal cases, that is, negative samples. Generally speaking, the proportion of abnormal cases in human judgment is much smaller than that of non-abnormal cases. The accuracy of the case abnormality recognition binary classifier obtained by data training is not high enough. Because the proportion of positive samples (that is, abnormal cases) in the trainin...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the embodiments of the present application is to propose a tree-structure-based abnormal case identification method, device, equipment, and storage medium to solve the existing technical problem of insufficient accuracy of the case identification model obtained through artificially labeled data training

Images