Anti-fraud module decision fusion method based on online transaction characteristics

Abstract

The invention provides an anti-fraud module decision fusion method based on online transaction characteristics. The anti-fraud module decision fusion method comprises the following steps: S1, settinga plurality of anti-fraud modules and characteristic data according to a specific scene; S2, dividing the training data set into a plurality of data subsets according to the feature data; S3, trainingeach anti-fraud module by using the plurality of data subsets; s4, calculating performance indexes of the anti-fraud modules respectively, and adjusting parameters of the anti-fraud modules accordingto the performance indexes to improve the performance of the anti-fraud modules; s5, arranging to obtain all logic combination modes of each anti-fraud module, and carrying out integrated training oneach logic combination mode to obtain an integrated model; and S6, judging whether the current transaction is a fraudulent transaction or not by utilizing the integrated model. According to the anti-fraud module decision fusion method based on the online transaction characteristics, the corresponding decision scheme can be adaptively generated according to the characteristics of the transaction data, and the judgment result about whether the transaction is fraud or not is obtained according to the decision scheme.

Description

technical field [0001] The invention relates to the technical field of Internet financial transaction anti-fraud, in particular to an anti-fraud module decision fusion method based on online transaction characteristics. Background technique [0002] With the development of social economy and Internet technology, the financial sector has gradually expanded its business to Internet finance. Internet finance has quickly been recognized by the public due to its convenience. However, due to the complexity of the Internet environment, both parties to transactions through the Internet cannot accurately assess the fraud risks of transactions, and fraud problems caused by identity theft and equipment loss often occur. Therefore, anti-fraud work against Internet finance is particularly important. [0003] At this stage, the massive amount of Internet financial transaction data makes the anti-fraud method based on machine learning a mainstream research direction. Use massive data to ...

AI Technical Summary

Problems solved by technology

[0011] To sum up, there are problems in the existing technology: the system has a single decision-making method, only uses a simple combination strategy for integration, relies too much on the independence between modules, is not sensitive to different types of fraudulent transactions, and has poor self-adaptability. Transaction data characteristics choose different decision-making methods

Images