WO2019192136A1

WO2019192136A1 - Electronic device, financial data processing method and system, and computer-readable storage medium

Info

Publication number: WO2019192136A1
Application number: PCT/CN2018/102226
Authority: WO
Inventors: 李正洋; 毛小豪
Original assignee: 平安科技（深圳）有限公司
Priority date: 2018-04-03
Filing date: 2018-08-24
Publication date: 2019-10-10
Also published as: CN108734335A

Abstract

Disclosed in the present application are an electronic device, a financial data processing method and system, and a computer-readable storage medium. According to the present application, a factor data prediction model corresponding to factor data to be processed is determined, a predicted value of said factor data is calculated according to the factor data prediction model, and the observed value and predicted value of the factor data within a first preset time interval are input to a data processing model to obtain denoised factor data. Compared with the prior art, the present application eliminates the noise in the observed value of the factor data, improves the accuracy of the factor data, and lays a good foundation for the subsequent analysis of the factor data.

Description

Electronic device, financial data processing method, system and computer readable storage medium

Priority claim

The present application is based on the priority of the Chinese Patent Application entitled "Electronic Device, Financial Data Processing Method, and Computer Readable Storage Media", filed on Apr. 3, 2018, the disclosure of which is incorporated herein by reference. The entire content is incorporated herein by reference.

Technical field

The present application relates to the field of computer technologies, and in particular, to an electronic device, a financial data processing method, a system, and a computer readable storage medium.

Background technique

Financial data includes stock factor data, bond factor data, futures factor data, fund factor data, and so on. For example, factor analysis based on stock factor data is often used as an important basis for judging the future trend of the stock. There are various types of factor data, and different types of factor data have different calculation methods. Due to the collection method or spatiotemporality, the observed factor data is basically considered to be unclean (ie, full of noise) data. For example, due to the delay, when the momentum factor observed at a certain time is positive, it does not mean that the stock price will continue to increase upward, and there may be momentum reversal. Therefore, the analysis results based on the data filled with noise are not accurate.

Summary of the invention

The main purpose of the present application is to solve the problem that the observed factor data is full of noise, and the factor analysis result based on the factor data is not accurate.

To achieve the above object, an electronic device proposed by the present application, the electronic device includes a memory and a processor, and the memory stores a financial data processing system operable on the processor, the financial data processing system being The following steps are implemented when the processor is executed:

S10. Obtain an observation value of each to-be-processed factor data in the first preset time interval.

S20, in a pre-established factor data prediction model library, querying a factor data prediction model corresponding to each of the to-be-processed factor data according to a mapping relationship between the predetermined factor data and the factor data prediction model;

S30. Acquire a predicted value of each of the to-be-processed factor data in the first preset time interval according to the factor data prediction model corresponding to each of the to-be-processed factor data.

S40. Input the acquired observation value of each of the to-be-processed factor data and the predicted value of each of the to-be-processed factor data into a predetermined data processing model to obtain the corrected factor data.

In addition, to achieve the above object, the present application provides a financial data processing method, the method comprising the steps of:

In addition, in order to achieve the above object, the present application provides a financial data processing system, where the financial data processing system includes:

a first acquiring module, configured to acquire an observation value of each to-be-processed factor data in the first preset time interval;

a query module, configured to query, in a pre-established factor data prediction model library, a factor data prediction model corresponding to each of the to-be-processed factor data according to a mapping relationship between the predetermined factor data and the factor data prediction model;

a second acquiring module, configured to acquire, according to the factor data prediction model corresponding to each of the to-be-processed factor data, a predicted value of each of the to-be-processed factor data in the first preset time interval;

And a data processing module, configured to input the acquired observation value of each of the to-be-processed factor data and the predicted value of each of the to-be-processed factor data into a predetermined data processing model to obtain the corrected factor data.

In addition, in order to achieve the above object, the present application provides a computer readable storage medium, wherein the computer readable storage medium stores a financial data processing system, and the financial data processing system can be executed by at least one processor. Taking the at least one processor to perform the following steps:

The present application determines a factor data prediction model corresponding to the to-be-processed factor data, calculates a predicted value of the to-be-processed factor data according to the factor data prediction model, and inputs the observed value and the predicted value of the factor data in the first preset time interval to the data. The model is processed to obtain the denoised factor data; compared with the prior art, the present application fully considers the difference in characteristics between the various factor data, and the prediction value of the factor data can be made by using different factor data prediction models for it. Accurate and more in line with the real situation. In addition, the factor data after denoising by the data processing model eliminates the noise in the observation data of the factor data, improves the accuracy of the factor data, and lays a good foundation for the subsequent analysis of the factor data.

DRAWINGS

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings to be used in the embodiments or the prior art description will be briefly described below. Obviously, the drawings in the following description are only It is a certain embodiment of the present application, and other drawings can be obtained according to the structures shown in the drawings without any creative work for those skilled in the art.

1 is a schematic flowchart of a first embodiment of a method for processing financial data according to the present application;

2 is a schematic flowchart of a second embodiment of a method for processing financial data according to the present application;

3 is a schematic diagram of an operating environment of a first embodiment of a financial data system of the present application;

4 is a block diagram of a program of a first embodiment of the financial data system of the present application;

FIG. 5 is a block diagram of a program of a second embodiment of the financial data system of the present application.

The implementation, functional features and advantages of the present application will be further described with reference to the accompanying drawings.

detailed description

The principles and features of the present application are described in the following with reference to the accompanying drawings, which are only used to explain the present application and are not intended to limit the scope of the application.

As shown in FIG. 1, FIG. 1 is a schematic flowchart of a first embodiment of a method for processing financial data according to the present application.

In this embodiment, the method includes:

Taking the factor data of the stock as an example, the above factor data may include the stock's day gain, moving average, stock trading volume, and the like.

The user can set the first preset time interval as needed, for example, setting the first preset time interval from January 2, 2018 to January 22, 2018.

The frequency of obtaining the observation value of each to-be-processed factor data in the first preset time interval may be set as needed, for example, in real time, or set a fixed time interval (for example, 1 day), or receiving an acquisition instruction issued by the user. The step of acquiring the observed values of the respective to-be-processed factor data in the first preset time interval is performed. It should be noted that if some factor data (for example, net profit growth, operating profit growth, etc.) needs to be calculated, the calculation step may be performed in advance, and the factor data obtained by the calculation step is stored to a preset storage space. When the step S10 is performed, the required factor data may be read in the storage space. In some embodiments, the step S10 may also include the calculation step of the factor data.

A corresponding factor data prediction model is established in advance for each factor data. Since the characteristics of each factor data are different, a method for establishing a corresponding factor data prediction model is also different, and the established factor data prediction models are stored in the factor data prediction model library. In the middle, the mapping relationship between the factor data and the factor data prediction model is saved (for example, a mapping table reflecting the mapping relationship between the factor data and the factor data prediction model).

For example, if the predicted value of the to-be-processed factor data at time K is to be obtained, the observation value of the to-be-processed factor data at time K-1 is input into the factor data prediction model, and the output result of the factor data prediction model is K-time waiting The predicted value of the processing factor data.

The embodiment determines a factor data prediction model corresponding to the to-be-processed factor data, calculates a predicted value of the to-be-processed factor data according to the factor data prediction model, and inputs the observed value and the predicted value of the factor data in the first preset time interval to The data processing model is used to obtain the denoised factor data; compared with the prior art, the present embodiment fully considers the characteristics of each factor data, and uses different factor data prediction models to predict the factor data. More accurate and more in line with the real situation. In addition, the factor data after denoising by the data processing model eliminates the noise in the observation of the factor data, improves the accuracy of the factor data, and lays a good foundation for the subsequent analysis of the factor data. basis.

Preferably, in this embodiment, the foregoing step S40 may be specifically:

Obtaining the obtained observation values of the to-be-processed factor data in chronological order to generate an observation matrix of the to-be-processed factor data;

And predicting the obtained predicted values of the to-be-processed factor data in a chronological order to generate a prediction matrix of the to-be-processed factor data;

The observation matrix of the to-be-processed factor data and the prediction matrix of the to-be-processed factor data are input to a predetermined Kalman filter model to obtain the corrected factor data.

In this embodiment, the step of generating the observation matrix of the to-be-processed factor data may be performed immediately after the step S10 is performed; likewise, the step of generating the prediction matrix of the to-be-processed factor data may be performed immediately after the step S30 is performed.

In this embodiment, the data processing model is a Kalman filter model, and the Kalman filter model can effectively eliminate the noise of the factor data, so that the factor data is closer to its true value.

As shown in FIG. 2, FIG. 2 is a schematic flowchart of a second embodiment of a method for processing financial data according to the present application.

In the second embodiment of the method for processing financial data of the present application, the present embodiment is based on the first embodiment. Before the step S20, the method further includes:

S50: classify each factor data according to a predetermined classification rule;

S60. The factor data prediction model corresponding to each of the factor data is respectively established according to the type of each of the factor data.

S70: Store the mapping relationship between each of the established factor data prediction models and their corresponding factor data.

Preferably, in this embodiment, the foregoing step S60 includes the following steps:

When the type of the one-factor data is the first data type, the factor data prediction model corresponding to the factor data is established based on the following operation expression:

X(K+1)=Z(K)

Here, X(K+1) is a predicted value of the factor data at time K+1, and Z(K) is an observed value of the factor data at time K.

The method for establishing a factor data prediction model corresponding to the factor data of the first data type is applied to low-frequency factor data, and the so-called low-frequency factor data is factor data that changes slowly with price.

Preferably, in the embodiment, the foregoing step S60 further includes the following steps:

When the type of the one-factor data is the second data type, the method for establishing the factor data prediction model corresponding to the factor data is:

Collecting the factor data in a second preset time interval;

Pre-processing (eg, normalization) of the collected factor data;

The pre-processed factor data is used, and a factor data prediction model corresponding to the factor data is constructed based on a long-term and short-term memory neural network.

The step of constructing the factor data prediction model corresponding to the factor data based on the Long Short-Term Memory (LSTM) neural network is specifically as follows:

Based on cross-validation, the pre-processed factor data is divided into a training set, an evaluation set, and a test set (eg, 70% of the factor data is used as the training set, and 10% of the factor data is used as the evaluation set, 20% of the factor data as a test set). The training set is input to the LSTM neural network model and trained based on a gradient descent method (eg, a stochastic gradient descent method). The evaluation set is used to verify the LSTM neural network model during the training process, and input the test set into the trained LSTM neural network model to verify the trained LSTM neural network model when training The obtained LSTM neural network model satisfies the first preset verification condition (for example, the difference from the verification result is less than a preset threshold), and the training is completed, and the trained LSTM neural network model is set as the factor data prediction of the factor data. model.

It should be noted that the above steps of dividing the factor data into the training set, the evaluation set and the test set based on the cross-validation method may be replaced by dividing the factor data into a training set and a test set based on the cross-validation method. And the number of training sets, evaluation sets, and test concentration factor data can be set as needed, and is not limited to the above-exemplified schemes.

The method for establishing a factor data prediction model corresponding to the factor data of the second data type is applicable to high-frequency factor data, and the so-called high-frequency factor data is factor data that changes rapidly with price, because the LSTM neural network targets high-frequency factors. The data has more accurate predictive power.

The method for judging whether a factor data belongs to low frequency factor data or high frequency factor data can be referred to the following example:

Variance method:

Taking the factor data in a time interval, and performing differential processing on the factor data with time as a variable;

Calculating a relative standard deviation of the differentially processed factor data as a variation frequency value;

If the change frequency value of the factor data is greater than or equal to the preset standard change frequency value, the factor data is determined as the high frequency factor data;

If the change frequency value of the factor data is less than the preset standard change frequency value, the factor data is determined as the low frequency factor data;

Wherein, if the stock factor data is taken as an example, the preset standard change frequency value may be set as a change frequency value of the closing price.

Intersection method:

The factor data in a time interval is taken, and the factor data is differentially processed by time to obtain a factor data curve. The coordinate X-axis of the factor data curve is the sampling time, and the Y-axis is the value of the differential processing factor data. ;

During the time interval, the number of intersections of the factor data curve and the mean line is counted; wherein the mean is an average of the factor data after the differential processing;

If the number of intersections of the one-factor data curve and the mean line is greater than or equal to the preset number of standard intersection points, the factor data corresponding to the factor data curve is determined as the high-frequency factor data;

If the number of intersections of the one-factor data curve and the mean line is less than the preset number of standard intersection points, the factor data corresponding to the factor data curve is determined as the low-frequency factor data;

Wherein, if the stock factor data is taken as an example, the preset number of standard intersection points may be set as the number of standard intersection points based on the closing price curve after the closing price differential processing.

In addition, in this embodiment, the above steps S50, S60 and S70 can be replaced by the following steps:

Collecting factor data in a third preset time interval;

Pre-processing the collected factor data;

The pre-processed factor data is divided into a training set and a test set according to a preset rule (for example, chronological order is used to divide 80% of the factor data as a training set, and divide 20% of the factor data as a test set) ;

Using the training set, and training based on multiple prediction models (for example, neural network model, random forest model, linear regression model, logistic regression model, etc.), using the test set to verify multiple prediction models of the completed training, according to predetermined The selection rule selects a prediction model from the plurality of prediction models as the factor data prediction model of the factor data.

The predetermined selection rule may be that the prediction model closest to the verification result is selected as the factor data prediction model.

The method for establishing the above-mentioned factor data prediction model is applicable to various factor data (this scheme does not need to classify the factor data), and simultaneously trains multiple prediction models based on one factor data to find an optimal prediction model as factor data of the factor data. Forecast model.

In addition, the present application also proposes a financial data processing system.

Please refer to FIG. 3 , which is a schematic diagram of an operating environment of a preferred embodiment of the financial data processing system 10 of the present application.

In the present embodiment, the financial data processing system 10 is installed and operated in the electronic device 1. The electronic device 1 may be a computing device such as a desktop computer, a notebook, a palmtop computer, and a server. The electronic device 1 may include, but is not limited to, a memory 11, a processor 12, and a display 13. Figure 3 shows only the electronic device 1 with components 11-13, but it should be understood that not all illustrated components may be implemented, and more or fewer components may be implemented instead.

The memory 11 may be an internal storage unit of the electronic device 1 in some embodiments, such as a hard disk or memory of the electronic device 1. The memory 11 may also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in hard disk equipped on the electronic device 1, a smart memory card (SMC), and a secure digital (SD). Card, flash card, etc. Further, the memory 11 may also include both an internal storage unit of the electronic device 1 and an external storage device. The memory 11 is used to store application software and various types of data installed in the electronic device 1, such as program codes of the financial data processing system 10. The memory 11 can also be used to temporarily store data that has been output or is about to be output.

The processor 12, in some embodiments, may be a Central Processing Unit (CPU), microprocessor or other data processing chip for running program code or processing data stored in the memory 11, such as performing financial data processing. System 10 and so on.

The display 13 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch sensor, or the like in some embodiments. The display 13 is for displaying information processed in the electronic device 1 and a user interface for displaying visualization. The components 11-13 of the electronic device 1 communicate with one another via a system bus.

Please refer to FIG. 4, which is a program block diagram of a preferred embodiment of the financial data processing system 10 of the present application. In the present embodiment, the financial data processing system 10 can be divided into one or more modules, one or more modules are stored in the memory 11, and by one or more processors (the processor 12 in this embodiment) Executed to complete the application. For example, in FIG. 4, the financial data processing system 10 can be divided into a first acquisition module 101, a query module 102, a second acquisition module 103, and a data processing module 104. A module referred to in this application refers to a series of computer program instruction segments capable of performing a specific function, and is more suitable than the program for describing the execution process of the financial data processing system 10 in the electronic device 1, wherein:

The first obtaining module 101 is configured to acquire an observation value of each to-be-processed factor data in the first preset time interval;

The query module 102 is configured to query, in a pre-established factor data prediction model library, a factor data prediction model corresponding to each of the to-be-processed factor data according to a mapping relationship between the predetermined factor data and the factor data prediction model;

a second obtaining module 103, configured to acquire, according to the factor data prediction model corresponding to each of the to-be-processed factor data, a predicted value of each of the to-be-processed factor data in the first preset time interval;

The data processing module 104 is configured to input the acquired observation value of each of the to-be-processed factor data and the predicted value of each of the to-be-processed factor data into a predetermined data processing model to obtain the corrected factor data.

The frequency of obtaining the observation value of each to-be-processed factor data in the first preset time interval may be set as needed, for example, in real time, or set a fixed time interval (for example, 1 day), or receiving an acquisition instruction issued by the user. The step of acquiring the observed values of the respective to-be-processed factor data in the first preset time interval is performed. It should be noted that if some factor data (for example, net profit growth, operating profit growth, etc.) needs to be calculated, the calculation step may be performed in advance, and the factor data obtained by the calculation step is stored to a preset storage space. When the step of acquiring the observed values of the to-be-processed factor data in the first preset time interval is performed, the first obtaining module 101 may read the required factor data in the storage space, and in some embodiments, the first obtaining Module 101 can also perform the calculation steps of the factor data.

Before the query module 102 performs the query step, the financial data processing system 10 pre-establishes a corresponding factor data prediction model for each factor data. Since the characteristics of each factor data are different, the method for establishing a corresponding factor data prediction model is different. The established factor data prediction models are stored in the factor data prediction model library, and the mapping relationship between the factor data and the factor data prediction model is saved (for example, a mapping table reflecting the mapping relationship between the factor data and the factor data prediction model).

The step of acquiring, by the second acquiring module 103, the predicted values of the to-be-processed factor data in the first preset time interval may be specifically exemplified by: if the predicted value of the to-be-processed factor data at the time K is obtained, then K is obtained. The observation value of the factor data to be processed at time -1 is input to the factor data prediction model, and the output result of the factor data prediction model is the predicted value of the factor data to be processed at time K.

Preferably, in the embodiment, the data processing module 104 is further configured to:

In this embodiment, the step of generating the observation matrix of the to-be-processed factor data may be performed by the first obtaining module 101 immediately after performing the step of acquiring the observed values of the to-be-processed factor data in the first preset time interval; Similarly, the step of generating the prediction matrix of the to-be-processed factor data may be performed by the second obtaining module 103 immediately after performing the step of acquiring the predicted values of the to-be-processed factor data in the first preset time interval. .

As shown in FIG. 5, FIG. 5 is a program module diagram of a second embodiment of the financial data processing system of the present application.

In the second embodiment of the financial data processing system of the present application, the present embodiment is based on the first embodiment, the financial data processing system 10 further includes a model building module 105, and the model building module 105 is configured to:

Performing classification processing on each factor data according to a predetermined classification rule; respectively, establishing a factor data prediction model corresponding to each of the factor data according to each type of the factor data; and correspondingly establishing each of the determined factor data prediction models The mapping relationship between the factor data is stored.

Preferably, in the embodiment, the model building module 105 is further configured to:

X(K+1)=Z(K)

Collecting the factor data in a second preset time interval;

Pre-processing (eg, normalization) of the collected factor data;

The method for establishing a factor data prediction model corresponding to the factor data of the second data type is applicable to high-frequency factor data, and the so-called high-frequency factor data is factor data that changes rapidly with price, because the LSTM neural network is for the time series Factor data has more accurate predictive power.

In addition, in this embodiment, the model building module 105 described above is further configured to:

Collecting factor data in a third preset time interval;

Pre-processing the collected factor data;

Further, the present application further provides a computer readable storage medium storing a financial data processing system, the financial data processing system being executable by at least one processor to cause the at least one processing The financial data processing method in any of the above embodiments is performed.

The above description is only a preferred embodiment of the present application, and is not intended to limit the scope of the patents of the present application, and the equivalent structural transformation, or direct/indirect use, of the present application and the contents of the drawings is used in the present invention. All other related technical fields are included in the patent protection scope of the present application.

Claims

An electronic device, comprising: a memory and a processor, the memory storing a financial data processing system operable on the processor, the financial data processing system being The following steps are implemented during execution:

S10. Obtain an observation value of each to-be-processed factor data in the first preset time interval.

S20, in a pre-established factor data prediction model library, querying a factor data prediction model corresponding to each of the to-be-processed factor data according to a mapping relationship between the predetermined factor data and the factor data prediction model;

S30. Acquire a predicted value of each of the to-be-processed factor data in the first preset time interval according to the factor data prediction model corresponding to each of the to-be-processed factor data.

S40. Input the acquired observation value of each of the to-be-processed factor data and the predicted value of each of the to-be-processed factor data into a predetermined data processing model to obtain the corrected factor data.
The electronic device according to claim 1, wherein the processor is further configured to execute the financial data processing system to perform the following steps before the step S20:

S50: classify each factor data according to a predetermined classification rule;

S60. The factor data prediction model corresponding to each of the factor data is respectively established according to the type of each of the factor data.

S70: Store the mapping relationship between each of the established factor data prediction models and their corresponding factor data.
The electronic device of claim 2, wherein the step S60 comprises:

When the type of the one-factor data is the first data type, the factor data prediction model corresponding to the factor data is established based on the following operation expression:

X(K+1)=Z(K)

Here, X(K+1) is a predicted value of the factor data at time K+1, and Z(K) is an observed value of the factor data at time K.
The electronic device of claim 2, wherein the step S60 further comprises:

When the type of the one-factor data is the second data type, the method for establishing the factor data prediction model corresponding to the factor data includes:

Collecting the factor data in a second preset time interval;

Pre-processing the collected factor data;

The pre-processed factor data is used, and a factor data prediction model corresponding to the factor data is constructed based on a long-term and short-term memory neural network.
The electronic device according to any one of claims 1 to 4, wherein the step S40 comprises:

Obtaining the obtained observation values of the to-be-processed factor data in chronological order to generate an observation matrix of the to-be-processed factor data;

And predicting the obtained predicted values of the to-be-processed factor data in a chronological order to generate a prediction matrix of the to-be-processed factor data;

The observation matrix of the to-be-processed factor data and the prediction matrix of the to-be-processed factor data are input to a predetermined Kalman filter model to obtain the corrected factor data.
A financial data processing method, characterized in that the method comprises the steps of:

S10. Obtain an observation value of each to-be-processed factor data in the first preset time interval.

S20, in a pre-established factor data prediction model library, querying a factor data prediction model corresponding to each of the to-be-processed factor data according to a mapping relationship between the predetermined factor data and the factor data prediction model;

S30. Acquire a predicted value of each of the to-be-processed factor data in the first preset time interval according to the factor data prediction model corresponding to each of the to-be-processed factor data.

S40. Input the acquired observation value of each of the to-be-processed factor data and the predicted value of each of the to-be-processed factor data into a predetermined data processing model to obtain the corrected factor data.
The financial data processing method according to claim 6, wherein before the step S20, the financial data processing method further comprises:

S50: classify each factor data according to a predetermined classification rule;

S60. The factor data prediction model corresponding to each of the factor data is respectively established according to the type of each of the factor data.

S70: Store the mapping relationship between each of the established factor data prediction models and their corresponding factor data.
The financial data processing method according to claim 7, wherein the step S60 comprises:

When the type of the one-factor data is the first data type, the factor data prediction model corresponding to the factor data is established based on the following operation expression:

X(K+1)=Z(K)

Here, X(K+1) is a predicted value of the factor data at time K+1, and Z(K) is an observed value of the factor data at time K.
The financial data processing method according to claim 7, wherein the step S60 further comprises:

When the type of the one-factor data is the second data type, the method for establishing the factor data prediction model corresponding to the factor data includes:

Collecting the factor data in a second preset time interval;

Pre-processing the collected factor data;

The pre-processed factor data is used, and a factor data prediction model corresponding to the factor data is constructed based on a long-term and short-term memory neural network.
The financial data processing method according to any one of claims 6 to 9, wherein the step S40 comprises:

Obtaining the obtained observation values of the to-be-processed factor data in chronological order to generate an observation matrix of the to-be-processed factor data;

And predicting the obtained predicted values of the to-be-processed factor data in a chronological order to generate a prediction matrix of the to-be-processed factor data;

The observation matrix of the to-be-processed factor data and the prediction matrix of the to-be-processed factor data are input to a predetermined Kalman filter model to obtain the corrected factor data.
A financial data processing system, characterized in that the financial data processing system comprises:

a first acquiring module, configured to acquire an observation value of each to-be-processed factor data in the first preset time interval;

a query module, configured to query, in a pre-established factor data prediction model library, a factor data prediction model corresponding to each of the to-be-processed factor data according to a mapping relationship between the predetermined factor data and the factor data prediction model;

a second acquiring module, configured to acquire, according to the factor data prediction model corresponding to each of the to-be-processed factor data, a predicted value of each of the to-be-processed factor data in the first preset time interval;

And a data processing module, configured to input the acquired observation value of each of the to-be-processed factor data and the predicted value of each of the to-be-processed factor data into a predetermined data processing model to obtain the corrected factor data.
The financial data processing system of claim 11 wherein said financial data processing system further comprises a model building module, said model building module for:

Sorting each factor data according to a predetermined classification rule;

Generating a factor data prediction model corresponding to each of the factor data according to each type of the factor data;

The mapping relationship between each of the established factor data prediction models and their corresponding factor data is stored.
The financial data processing system of claim 12, wherein the model building module is further configured to:

When the type of the one-factor data is the first data type, the factor data prediction model corresponding to the factor data is established based on the following operation expression:

X(K+1)=Z(K)

Here, X(K+1) is a predicted value of the factor data at time K+1, and Z(K) is an observed value of the factor data at time K.
The financial data processing system of claim 12, wherein the model building module is further configured to:

When the type of the one-factor data is the second data type, the method for establishing the factor data prediction model corresponding to the factor data includes:

Collecting the factor data in a second preset time interval;

Pre-processing the collected factor data;

The pre-processed factor data is used, and a factor data prediction model corresponding to the factor data is constructed based on a long-term and short-term memory neural network.
The financial data processing system according to any one of claims 11 to 14, wherein the data processing module is specifically configured to:

Obtaining the obtained observation values of the to-be-processed factor data in chronological order to generate an observation matrix of the to-be-processed factor data;

And predicting the obtained predicted values of the to-be-processed factor data in a chronological order to generate a prediction matrix of the to-be-processed factor data;

The observation matrix of the to-be-processed factor data and the prediction matrix of the to-be-processed factor data are input to a predetermined Kalman filter model to obtain the corrected factor data.
A computer readable storage medium, characterized in that the computer readable storage medium stores a financial data processing system, the financial data processing system being executable by at least one processor to cause the at least one processor to perform the following step:

S10. Obtain an observation value of each to-be-processed factor data in the first preset time interval.

S20, in a pre-established factor data prediction model library, querying a factor data prediction model corresponding to each of the to-be-processed factor data according to a mapping relationship between the predetermined factor data and the factor data prediction model;

S30. Acquire a predicted value of each of the to-be-processed factor data in the first preset time interval according to the factor data prediction model corresponding to each of the to-be-processed factor data.

S40. Input the acquired observation value of each of the to-be-processed factor data and the predicted value of each of the to-be-processed factor data into a predetermined data processing model to obtain the corrected factor data.
The computer readable storage medium of claim 16, wherein prior to said step S20, said processor is further configured to execute said financial data processing system to implement the following steps:

S50: classify each factor data according to a predetermined classification rule;

S60. The factor data prediction model corresponding to each of the factor data is respectively established according to the type of each of the factor data.

S70: Store the mapping relationship between each of the established factor data prediction models and their corresponding factor data.
The computer readable storage medium of claim 17, wherein the step S60 comprises:

When the type of the one-factor data is the first data type, the factor data prediction model corresponding to the factor data is established based on the following operation expression:

X(K+1)=Z(K)

Here, X(K+1) is a predicted value of the factor data at time K+1, and Z(K) is an observed value of the factor data at time K.
The computer readable storage medium of claim 17, wherein the step S60 further comprises:

When the type of the one-factor data is the second data type, the method for establishing the factor data prediction model corresponding to the factor data includes:

Collecting the factor data in a second preset time interval;

Pre-processing the collected factor data;

The pre-processed factor data is used, and a factor data prediction model corresponding to the factor data is constructed based on a long-term and short-term memory neural network.
A computer readable storage medium according to any one of claims 16 to 19, wherein said step S40 comprises:

Obtaining the obtained observation values of the to-be-processed factor data in chronological order to generate an observation matrix of the to-be-processed factor data;

And predicting the obtained predicted values of the to-be-processed factor data in a chronological order to generate a prediction matrix of the to-be-processed factor data;

The observation matrix of the to-be-processed factor data and the prediction matrix of the to-be-processed factor data are input to a predetermined Kalman filter model to obtain the corrected factor data.