Financial risk prediction method and device based on migration sample screening and electronic equipment

Abstract

The invention provides a financial risk prediction method and device based on migration sample screening and electronic equipment. The method comprises the following steps: acquiring historical sampledata, and determining a source domain training set and a target domain training set; training a plurality of sub-models based on the target domain training set, and selecting at least two sub-modelsfrom the plurality of trained sub-models; performing migration sample screening in the source domain data set by using the at least two selected sub-models according to a sample screening strategy toform a migration sample training set; combining the formed migration sample training set with a target domain training set to obtain a new target domain training set; training a financial risk prediction model by using the new target domain data set; and calculating a user risk prediction value of the target user in the target domain by using the trained financial risk prediction model. Accordingto the method, the problems of high deviation and high variance caused by sample data can be effectively avoided, and data resources are effectively saved.

Description

technical field [0001] The invention relates to the field of computer information processing, in particular to a financial risk prediction method, device and electronic equipment based on migration sample screening. Background technique [0002] Risk control (referred to as risk control) means that risk managers take various measures and methods to eliminate or reduce the various possibilities of risk cases, or risk controllers reduce the losses caused by risk cases. Risk control is generally used in the financial industry, such as risk control for company transactions, business transactions or personal transactions. [0003] In the prior art, the main purpose of financial risk assessment is how to distinguish good customers from bad customers, evaluate the risk status of users, and reduce credit risk to maximize profits. However, many algorithms require a large amount of data when learning, especially in deep learning, because a large number of parameters need to be learne...

AI Technical Summary

Problems solved by technology

[0006] For the situation where there is no or only a small amount of data or the data distribution is uneven, how to establish a more effective model with higher model stability and model accuracy, the present invention provides a financial risk prediction method based on migration sample screening, including: Obtaining historical sample data, determining a source domain training set and a target domain training set; training multiple sub-models based on the target domain training set, and selecting at least two sub-models from the trained multiple sub-models; using the selected at least The two sub-models perform migration sample screening in the source domain data set according to the sample screening strategy to form a migration sample training set; merge the formed migration sample training set with the target domain training set to obtain a new target domain training set ; Use the new target domain data set to train the financial risk prediction model; use the trained financial risk prediction model to calculate the user risk prediction value of the target user on the target domain

Images