CN-122023036-A - Financial risk monitoring method

CN122023036ACN 122023036 ACN122023036 ACN 122023036ACN-122023036-A

Abstract

The application discloses a financial risk monitoring method. The method comprises the steps of collecting financial data of each service system in real time, adopting a CNN-LSTM-Attention mixed neural network model to conduct calculation analysis on the financial data to obtain calculation analysis results, conducting multidimensional matching on the calculation analysis results, and generating early warning information according to the matching results. According to the application, the CNN-LSTM-attribute mixed neural network model is adopted to perform real-time dynamic calculation and analysis on the acquired financial data, so that the risk monitoring effect is improved.

Inventors

LI YIJIANG
CUI HAONAN
LIU HUI
Yang Tenghao
WANG XUEYI
Fu Junsha
JIANG FENG
CHENG ZHISONG
LI BO
JI PEIXI
DU YAN
WU DONGQI
ZHAO YIYANG

Assignees

人保信息科技有限公司
中国人民保险集团股份有限公司

Dates

Publication Date: 20260512
Application Date: 20251231

Claims (10)

1. A method of financial risk monitoring comprising: Collecting financial data of each business system in real time; adopting a CNN-LSTM-attribute mixed neural network model to calculate and analyze the financial data to obtain a calculation and analysis result; and carrying out multidimensional matching on the calculation analysis result, and generating early warning information according to the matching result.
2. The method of claim 1, wherein the computing the financial data using a CNN-LSTM-Attention hybrid neural network model to obtain a computed analysis result comprises: extracting local features of the financial data by adopting a convolutional neural network CNN layer in the CNN-LSTM-attribute mixed neural network model; capturing a time sequence dependency relationship between the local features by adopting a long-short-term memory network LSTM layer in the CNN-LSTM-attribute mixed neural network model; Adopting an Attention mechanism Attention layer in the CNN-LSTM-Attention mixed neural network model to carry out self-adaptive weighting on key time points on the time sequence dependency relationship so as to obtain weighted characteristic data; and obtaining the calculation analysis result based on the weighted characteristic data.
3. The method of claim 2, wherein the CNN layer comprises a three-layer CNN structure with convolution kernel sizes of 3 x 3,5 x 5, and 7 x 7, respectively, for extracting short-, medium-, and long-term local features of the financial data, respectively; The LSTM layer adopts a bidirectional LSTM structure, and the number of hidden units is 128, and the hidden units are respectively used for capturing the forward time sequence dependency relationship and the reverse time sequence dependency relationship between the local features; the Attention layer adopts an 8-head multi-head Attention mechanism.
4. A method according to claim 3, wherein said deriving said computational analysis results based on said weighted feature data comprises: carrying out feature fusion on the weighted feature data by adopting a multi-scale feature fusion algorithm based on wavelet packet decomposition to obtain fused feature data; adopting a financial index association graph construction algorithm to construct a dynamic association graph according to the fused characteristic data; And identifying the conduction path and the influence intensity of the financial risk according to the dynamic association diagram by adopting a causal reasoning-based risk conduction analysis algorithm.
5. The method according to claim 1, wherein the performing multidimensional matching on the calculation analysis result and generating the early warning information according to the matching result includes: Performing risk scoring on the calculation analysis result by adopting a four-dimensional early warning index system to obtain a risk scoring result, wherein the four dimensions comprise a financial dimension, a technical dimension, an environmental dimension and a compliance dimension; And generating the early warning information according to the comparison result of the risk scoring result and the risk scoring threshold value.
6. The method as recited in claim 5, further comprising: Based on a reinforcement learning mechanism, the risk scoring threshold is automatically adjusted according to the historical early warning effect and the business feedback.
7. The method as recited in claim 1, further comprising: and visually displaying the early warning information and/or the financial data in a plurality of terminals.
8. The method of claim 1, wherein after collecting the financial data of each business system in real time, further comprising: Performing fusion verification on the financial data acquired by each business system; performing format conversion on the financial data passing verification; and carrying out real-time data stream processing on the financial data after format conversion to prevent data extrusion.
9. The method of claim 1, wherein after collecting the financial data of each business system in real time, further comprising: storing the time sequence data in the financial data into a distributed time sequence database; And storing the document data in the financial data into a document database.
10. The method of claim 9, wherein storing the timing data in the financial data into a distributed timing database comprises: analyzing the time locality and the space locality of the time sequence data, and carrying out slicing division on the time sequence data by adopting a clustering algorithm to obtain slicing time sequence data; Carrying out consistency processing on each piece of time sequence data by adopting a consensus algorithm to obtain consensus data; And generating cache data according to the consensus data, and storing the cache data into the distributed time sequence database.

Description

Financial risk monitoring method Technical Field The application belongs to the technical field of financial risk management, and particularly relates to a financial risk monitoring method. Background With the penetration of enterprise digital transformation, the financial risk monitoring has evolved from traditional post-supervision to real-time monitoring and predictive early warning. In the related art, a traditional financial management system represented by an SAP FI module is mainly adopted, but the analysis capability of the traditional financial management system is limited to a predefined analysis report, the dynamic analysis capability is lacked, and the risk monitoring effect is poor. Disclosure of Invention The embodiment of the application aims to provide a financial risk monitoring method for solving the problem of poor risk monitoring effect in the related art. In order to achieve the above purpose, the embodiment of the present application adopts the following technical scheme: in a first aspect, an embodiment of the present application provides a financial risk monitoring method, including collecting financial data of each service system in real time, performing calculation analysis on the financial data by using a CNN-LSTM-Attention hybrid neural network model to obtain a calculation analysis result, performing multidimensional matching on the calculation analysis result, and generating early warning information according to the matching result. The above at least one technical scheme adopted by the embodiment of the application can achieve the following beneficial effects: according to the embodiment of the application, financial data of each business system are collected in real time, the collected financial data are calculated and analyzed by adopting a CNN-LSTM-Attention mixed neural network model, a calculation and analysis result is obtained, multidimensional matching is carried out on the calculation and analysis result, and early warning information is generated according to the matching result. According to the embodiment of the application, the CNN-LSTM-Attention mixed neural network model is adopted to perform real-time dynamic calculation and analysis on the acquired financial data, so that the risk monitoring effect is improved. Drawings The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute a limitation on the application. In the drawings: FIG. 1 is a flow chart of a method for monitoring financial risk according to an embodiment of the present application; FIG. 2 is a flow chart of a method for monitoring financial risk according to another embodiment of the present application; FIG. 3 is a flow chart of a method for monitoring financial risk according to another embodiment of the present application; FIG. 4 is a flow chart of a method for monitoring financial risk according to another embodiment of the present application; FIG. 5 is a block diagram of a financial risk monitoring system according to one embodiment of the present application; FIG. 6 is a workflow diagram of a financial risk monitoring system provided by one embodiment of the present application; FIG. 7 is a user interface design of a financial risk monitoring system provided in accordance with one embodiment of the present application; FIG. 8 is a data flow processing schematic diagram of a financial risk monitoring system according to an embodiment of the present application; fig. 9 is a schematic structural diagram of an electronic device according to an embodiment of the present application. Detailed Description In order to make the objects, technical solutions and advantages of the present application more apparent, the technical solutions of the present application will be clearly and completely described below with reference to specific embodiments of the present application and corresponding drawings. It will be apparent that the described embodiments are only some, but not all, embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application. The terms "first," "second," and the like in this specification are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that embodiments of the application may be practiced otherwise than as specifically illustrated or described herein. Further, "and/or" in the present application means at least one of the connected objects, and the character "/" generally means a relationship in which the