Taxpayer credit evaluation method based on distributed automatic feature combination

Abstract

The invention discloses a taxpayer credit assessment method based on distributed automatic feature combination. The method comprises the following steps: 1) training a random forest model by using a training sample and adopting a MapReduce distributed computing framework to obtain a distributed random forest model; 2) inputting the training samples into the distributed random forest model, and generating a plurality of input combined features of each training sample; 3) combining the generated combined features with the feature information of the corresponding taxpayers; 4) training a score card model by using the combined features; and 5) for a taxpayer to be subjected to credit assessment, generating a combined feature of the taxpayer by using the distributed random forest model, combining the combined feature with the feature information of the taxpayer, and then inputting the combined feature of the taxpayer into the trained score card model to predict the credit score of the taxpayer. According to the invention, accurate credit assessment of taxpayers can be carried out.

Description

technical field [0001] The invention relates to a credit evaluation model and a taxpayer credit evaluation method, specifically a credit evaluation model and a taxpayer credit evaluation method for automatic feature combination through a distributed random forest, belonging to the field of computer big data processing. [0002] technical background [0003] Credit evaluation has been developed in the field of bank credit for decades. It is mainly used to evaluate the personal credit of applicants for loans, and assists the issuance of loans through credit evaluation, reducing the bank's capital income and risk of capital recovery. [0004] Taxpayer credit assessment in the tax field has only emerged in recent years, and it is mainly based on expert experience. Tax experts select indicators that can represent tax risks based on their professional experience, and assign different scores to different indicators. For each taxpayer, a lot of manual analysis and investigation are r...

AI Technical Summary

Problems solved by technology

However, the scorecard model can only process processed features. To obtain a more accurate credit scoring effect, a large number of professionals are required to construct carefully calculated indicators.

Images