Multi-source heterogeneous data fusion architecture system based on additive model

Abstract

The invention provides a multi-source heterogeneous data fusion architecture system based on an additive model. The multi-source heterogeneous data fusion architecture system comprises a multi-sourceheterogeneous data preprocessing module, a multi-source heterogeneous data fusion module and a model summing module. When the architecture system works, by calling a matched meta-model from a pre-established meta-model base, data of different formats can be processed, and data fusion is achieved through a model summing part. By means of the multi-source heterogeneous data fusion architecture system based on the additive model, the problem that during a multi-source heterogeneous data processing process, data easily gets lost, and needs to be matched with a target data source format according to the format of a data source can be effectively solved; meanwhile, the contribution of each newly added external data source to a target domain problem can also be analyzed, and high-speed processingand analyzing of mass heterogeneous data can be achieved.

Description

technical field [0001] The invention relates to a multi-source heterogeneous data processing system, in particular to a multi-source heterogeneous data fusion architecture system based on an additive model in the field of financial risk control, which belongs to the field of financial risk control big data processing. Background technique [0002] At present, with the development of my country's financial industry, the coverage of financial services is gradually expanding. For customers with loan needs, banks usually inquire about their previous loan records and the credit records of the borrower stored by the central bank, such as the loan amount, number of times, and whether they are repaid on time, etc. Occupation, social security information, income, expenses and other data are used to analyze the repayment ability of the borrower. With the development of the financial industry, especially the rapid popularization of Internet finance, the collection of information about...

AI Technical Summary

Problems solved by technology

For example, the age and income of the borrower are usually difficult to judge and compare

Because multi-source data presents the characteristics of low-level feature heterogeneity and high-level semantic correlation, traditional single-source learning methods can no longer adapt to the characteristics of multi-source data. Therefore, some processing methods for multi-source heterogeneous data have emerged.

[0004] However, the existing multi-source data integration is mainly to establish a bridge between the external source data and the target domain, to construct the transformation relationship between the external source data and the target domain, and then to solve the modeling problem of the target domain

This approach has three main disadvantages: 1. Information will be lost in the process of building bridges; 2. Generally, it only solves the problem of joint modeling between two domains, and cannot automatically integrate multiple data sources for the problem of the target domain. Explain that it is difficult to measure the contribution of newly added external data to the problem in the target field under given conditions. In the existing technology, the method of converting different data source formats first, packaging and modeling as a whole, cannot be judged Contribution value of newly added data sources to the results; 3. There are strict requirements on the format and generation logic of data sources in different fields. Adapt the format

It can realize the integration of various professional power grid parameters, and can calculate the parameter difference between each professional power grid parameters, and perform quality analysis and parameter correction on the parameters. Although this method can improve the accuracy and credibility of the analysis results, it cannot Analyze the contribution of each parameter to the overall

Images