A kind of risk assessment method and device

Abstract

The present invention provides a risk assessment method and device. The method includes: grouping data sources according to the risk information amount of the data, and obtaining a strongly correlated variable group including strongly correlated variables and a weakly correlated variable group including weak correlated variables; The weakly correlated variable group constructs a first risk assessment model; the first risk assessment model is used to perform prediction probability classification on the full sample containing only the weakly correlated variable, and a sample group with the highest proportion of negative samples is obtained; The sample group with the highest proportion of samples and the strongly correlated variable group construct a second risk assessment model; the first risk assessment model and the second risk assessment model are used for risk assessment. The invention solves the problem that the model prediction efficiency is not high due to the imbalance of positive and negative samples, and improves the prediction efficiency of the risk assessment model.

Description

technical field [0001] The invention relates to the technical field of risk control, and more particularly, to a risk assessment method and device. Background technique [0002] Risk assessment is the quantification of risk and a key technology in risk management. At present, risk assessment is generally carried out by modeling. In the process of model establishment, there are mainly steps such as data extraction, feature generation, feature selection, algorithm model generation and rationality evaluation. [0003] As the sources of data become more and more abundant, there are more and more data fields that can be used as risk characteristic variables. Since not all risk feature fields in all samples are valid values, the occurrence of missing values ​​is inevitable, and with the increase of feature fields, the missing situation will develop in a more and more serious direction. [0004] When the data is generally sparse, that is, there are many vacancies in the risk feat...

AI Technical Summary

Problems solved by technology

[0004] When the data is generally sparse, that is, when there are many vacant values ​​in the risk feature field, if the feature selection and subsequent modeling process are carried out according to the traditional model method, the efficiency of the model prediction is low. When using the model for risk assessment, the risk assessment is accurate. lower sex

Images