Auto finance risk control system with structured and unstructured data

Abstract

The invention discloses an auto financial risk control system containing structured and unstructured data. The implementation steps of the auto financial risk control system containing structured and unstructured data are as follows: data preprocessing, word segmentation, and n-grams algorithm , PCA algorithm, training, PCA reverse algorithm, set operation, loop, merge, the design of the auto financial risk control system containing structured and unstructured data is reasonable, and can effectively increase the number of features input by the risk control model, so that it can Directly improve the accuracy of risk control model prediction.

Description

technical field [0001] The invention is an auto finance risk control system including structured and unstructured data, belonging to the technical field of auto finance risk control. Background technique [0002] The business system of auto finance usually contains structured data and unstructured data. A common example of structured data is a table in a relational database. A column of a table represents a feature, and a row of the table represents a sample. At the same time, the automobile A large amount of unstructured data will be generated in the financial business process, such as the text information input by the salesman in the business process. The unstructured data contains a lot of information, and some information may not be reflected in the structured data. We Unstructured data can be mined and corresponding features can be extracted, and useful information in unstructured data can be added to the risk control model. [0003] Obviously, unstructured data has no...

AI Technical Summary

Problems solved by technology

[0003] Obviously, unstructured data has no fixed structure. For example, a customer order may contain dozens of words of text, while another order may contain hundreds of words of text, that is to say, different In the order, the length of the text is inconsistent. Our first goal is to use an algorithm to convert unstructured data into structured data, so that it can be merged with the original structured data and input into the model. However, when the text data is converted When it is structured data, tens of thousands or even hundreds of thousands of features may be generated, while a common auto finance business database only has thousands to tens of thousands of orders, so all the features converted into text are input into the model is impractical, we need a better way

Images