WO2020199483A1

WO2020199483A1 - Image processing method and apparatus for financial data, and device and computer-readable storage medium

Info

Publication number: WO2020199483A1
Application number: PCT/CN2019/103244
Authority: WO
Inventors: 李海疆
Original assignee: 平安科技（深圳）有限公司
Priority date: 2019-04-04
Filing date: 2019-08-29
Publication date: 2020-10-08
Also published as: CN110120082B; CN110120082A

Abstract

Disclosed are an image processing method and apparatus for financial data, and a computer device and a computer-readable storage medium. The method comprises: acquiring first data Kd of a day level and second data Kmin of a preset minute level (110), wherein the first data Kd is of a multi-dimensional data type, and the dimension thereof is higher than that of the second data Kmin; performing data processing on the first data Kd of the day level and the second data Kmin of the preset minute level, and establishing a vector autoregressive model of an image sample space according to the processed data (120); and inputting financial data to be processed into the vector autoregressive model, and after inverse transformation and decoding operations are performed, obtaining a prediction pattern of the financial data to be processed (130).

Description

Financial data image processing method, device, equipment and computer readable storage medium

Cross references to related applications

This application affirms that it enjoys the priority of the Chinese patent application filed on April 4, 2019 with the application number CN201910270950.9 and titled "Financial Data Image Processing Method, Device, Equipment and Readable Storage Medium". The Chinese patent application The overall content of is incorporated in this application by reference.

Technical field

The embodiments of the application relate to the technical field of financial data processing, and in particular to an image processing method, device, device, and computer-readable storage medium for financial data.

Background technique

The development of the modern economy has driven the growth of the financial market. All kinds of financial data in the financial market reflect all kinds of information in the financial market. Financial data, including futures, stocks, interest rates, and credit, can reflect the current financial market. State, and drawing financial data into a change curve image or table can show this state more intuitively, which can be used as a reference for the next transaction behavior.

The inventor realizes that the current financial images of the forecast type focus on the return rate and the relevant indicators derived from the return rate, and use a certain machine learning algorithm on the basis of the factor model to predict the future financial data. However, since this method is limited to the rate of return and its related indicators, this makes this processing prediction method have certain limitations, and it is impossible to make more accurate predictions.

There is such a theory, the law of manifold distribution, which can be understood as: the same type of high-dimensional data in nature is often concentrated near a low-dimensional manifold. Regarding the manifold distribution law, the current theoretical development is not complete, and often the effect of machine learning lies in parameter tuning. However, many practical application problems can be modeled in the framework of manifolds, which can be described and sorted out in geometric language, and solved with geometric theoretical tools. For financial data images, for the most common K-line chart, investors are most concerned about predicting the trend of the future K-line chart. However, there is no corresponding processing method for the law of manifold distribution applied to the processing of financial data.

Summary of the invention

In order to overcome the problems in the related technology, this application provides an image processing method, device, computer equipment, and computer-readable storage medium for financial data, so as to achieve a more reliable and accurate preset of financial data trends from the perspective of image recognition.

In the first aspect, an embodiment of the present application provides an image processing method for financial data, the method including:

Acquire first data Kd of daily level and second data Kmin of preset minute level, where the first data Kd is a multidimensional data type and has a higher dimension than the second data Kmin;

Data processing is performed on the first data Kd of the daily level and the second data Kmin of the preset minute level, and a vector autoregressive model of the image sample space is established according to the processed data;

The financial data to be processed is input into the vector autoregressive model to obtain a prediction pattern of the financial data to be processed after inverse transformation and decoding operations.

In the second aspect of this application, this application also relates to an image processing device for financial data, the device comprising:

The obtaining module is used to obtain the first data Kd at the daily level and the second data Kmin at the preset minute level, wherein the first data Kd is a multidimensional data type and has a higher dimension than the second data Kmin;

A processing module, configured to perform data processing on the first data Kd of the daily level and the second data Kmin of the preset minute level, and establish a vector autoregressive model of the image sample space according to the processed data;

The regression module is used to input the financial data to be processed into the vector autoregressive model to obtain the predicted pattern of the financial data to be processed after inverse transformation and decoding operations.

In a third aspect, the present application also provides a computer device, including a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the following steps when the processor executes the computer program:

In a fourth aspect, this application also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

This application constructs a vector autoregressive model by establishing a homeomorphic mapping from the image sample space to the feature space on the K-line image samples of financial data based on the law of manifold distribution. When the obtained financial data is input into the model, the financial data outside the sample After the data, the image display of future financial data can be obtained through the calculation of the model, and the financial data can be predicted from the perspective of image recognition, which replaces the traditional method of predicting the rate of return and can be extended to all other types of K Line graph data has a wide range of applications.

Description of the drawings

Fig. 1 is a schematic flowchart showing a method for image processing of financial data according to an exemplary embodiment.

Fig. 2 is a block diagram showing an image processing device for financial data according to an exemplary embodiment.

Fig. 3 is a block diagram showing a computer device for implementing a method according to an exemplary embodiment.

detailed description

The application will be further described in detail below with reference to the drawings and embodiments. It is understandable that the specific embodiments described here are only used to explain the application, but not to limit the application. In addition, it should be noted that, for ease of description, the drawings only show a part of the structure related to the present application instead of all of the structure.

Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the steps are described as sequential processing in the flowchart, many of the steps can be implemented in parallel, concurrently, or simultaneously. In addition, the order of the steps may be rearranged, and the process may be terminated when the operation is completed, but there may also be other steps not included in the drawings. Processing can correspond to methods, functions, procedures, subroutines, subroutines, and so on.

This application relates to an image processing method, device, computer equipment, and computer readable storage medium for financial data. It is mainly used in scenarios where financial data is predicted. The basic idea is: the calculation of financial data based on the law of manifold distribution K-line image samples establish a homeomorphic mapping from the image sample space to the feature space, and construct a vector autoregressive model. When the financial data obtained, that is, the financial data outside the sample, is input into the model, the future results can be obtained through the calculation of the model The image display of financial data realizes the prediction of financial data from the perspective of image recognition, which replaces the traditional method of forecasting with yield indicators, and can be extended to all other types of K-line graph data, with a wide range of applications.

This embodiment is applicable to the case of an intelligent terminal with a vector autoregressive model for image processing of financial data. The method can be executed by a control device of the vector autoregressive model, where the device can be software and/ Or hardware implementation, generally integrated in a smart terminal, or controlled by a central control module in the terminal, as shown in Figure 1, which is a schematic diagram of the basic flow of a financial data image processing method of this application. Specifically include the following steps:

In step 110, first data Kd at a daily level and second data Kmin at a preset minute level are acquired, where the first data Kd is a multidimensional data type and has a higher dimension than the second data Kmin;

The daily level may be 1 trading day level, and the preset minute level may be 5 minutes as the preset level. In a feasible implementation manner of this application, taking the Shanghai and Shenzhen 300 Index as an example, first extract 100 consecutive 100 The daily data of each trading day, the data of each trading day includes K _d equal to the first data Kd: (open price, daily opening price, close price, daily closing price, high price, daily highest price, low price, the lowest daily price), and K _5min equivalent to the second data Kmin: all open and close prices at the 5-minute level in the trading day.

The first data Kd has multiple data types, for example, financial data in four dimensions including open price, close price, high price, and low price, and the second data Kmin includes two dimensions of open price, close price, etc. Data, and generally the dimension of the data type of the first data Kd is higher than that of the second data Kmin.

In the exemplary embodiment of the present application, in terms of A shares, the trading time of each trading day is 4 hours, so n=48 first data Kd of daily level are included in one trading day.

In step 120, data processing is performed on the first data Kd of the daily level and the second data Kmin of the preset minute level, and a vector autoregressive model of the image sample space is established according to the processed data;

In a feasible implementation manner of the exemplary embodiment of the present application, the sample coordinate mapping is established by using the K-line chart sample with the A-share period of 5 days, then:

To

Represents the mapping sequence of coordinate mapping for a K-line chart sample with a period of 5 days,

Form a local coordinate card.

In another implementation of the exemplary embodiment of the present application, taking the Shanghai and Shenzhen 300 Index as an example, daily data of 100 consecutive trading days are extracted, and the data of each trading day includes K _d : (open price, close price, high price, low price), and K _5min : All 5-minute-level open price and close price in this trading day, the data processing process includes; for 5-minute-level open price and close price, calculate the arithmetic average value, namely (open price+close price)/2, called the 5-minute average, count the median of 48 5-minute averages within a trading day. The specific method is: sort the 48 5-minute averages from small to large, and take the order as The 24th and 25th data are added and divided by 2, which is the median M; calculate the intraday amplitude S=high price-low price; calculate the daily K-line chart entity size Pr=|open price-close price |, that is, the absolute value of the difference between the opening price and the closing price of the trading day. Therefore, for a K-line chart of a trading day, the relationship between the first data Kd of the daily level and the second data Kmin of the preset minute level and the K-line chart, the median M and the intraday amplitude S is established. Mapping:

This mapping is a comorphic mapping and enables the conversion of high-dimensional manifolds to low-dimensional manifolds.

The vector autoregressive model is an econometric model that is used to estimate the dynamic relationship of joint endogenous variables. It contains multiple sequence variables. The process of establishing a vector autoregressive model includes: establishing time series mapping and performing coordinate mapping on K-line chart samples ，Establish the time series of the three-dimensional vector (Pr,M,S) according to the above mapping: x _i =(Pr _i ,M _i ,S _i ), through the regression equation

x _n is the k-line chart on the nth day, c is the three-dimensional constant vector, ε is the error vector, and _Ai is the 3*3 dimensional regression coefficient matrix. Determine the desired vector data or mapping data.

_{_{Wherein, Pr i, M i, S}} i K are the physical size of the i-th line graph of days, and the median day amplitude.

In a feasible implementation of the exemplary embodiment of the present application, K _d represents the sample point of the daily K-line chart, including four data of the daily K-line chart: open price (day opening price), close price (day Closing price), high price (daily highest price), low price (daily lowest price);

K _5min represents the opening and closing prices of all 5-minute K-line charts in a trading day. It is a collection of two-dimensional data. Because the daily trading time of A shares is 4 hours, the sample capacity of a K _5min is 48 .

When the first data Kd of financial data for four consecutive trading days and K5min of the preset minute level are known, and the forecast pattern for the 5th day needs to be obtained, formula 1 for establishing the mapping includes:

i=1,2,3,4,5, where θ ₁ , θ ₂ , θ ₃ , θ ₄ represent the four-dimensional data of the daily level, and θ ₅ represents the median of the 5-minute level data within the day, K _d represents the sample point of the daily K-line chart, K _5min represents the intra-day 5-minute K-line chart in a trading day,

Represents the mapping sequence of coordinate mapping for a K-line chart sample with a period of 5 days, i = 1, 2, 3, 4, 5,

Form a local coordinate card.

The process of constructing a vector autoregressive model (VAR model) includes: when the first data Kd of financial data for four consecutive trading days and the preset minute level K5min are known, the forecast pattern of the 5th day needs to be obtained, and the regression equation

Where n=5, as follows:

Among them, x _n is the k-line graph on the nth day, c is a three-dimensional constant vector, ε is an error vector, and A _i is a 3*3 dimensional regression coefficient matrix. Through the vector autoregressive model, A _i and c can be obtained, namely Regression obtains 3*3 dimensional regression coefficient matrix and 3D constant vector.

In the above regression equation of the exemplary embodiment of the present application, the regression coefficient matrix can be solved in the following logical sequence: the data vector on the fifth day is regressed with respect to the data vector on the first to fourth days, and the vector on the sixth day is related to the second The vector regression from day to day 5, and so on, is carried out on a rolling basis, a total of 96 times. The optimal solution of the regression is formula 2:

x _i is the sample point, which is an n*m-dimensional matrix in this scheme. The L2 norm of the matrix is defined as the square root of the maximum eigenvalue of ^x _i ^T x _i

The purpose of the vector autoregressive model is to find the mapping

Make

{x _i }→R ^d , R ^d is a d-dimensional vector space, {x _i } is a set of sample points.

At this point, the loss function

The loss function is minimized.

In step 130, input the obtained financial data into the vector autoregressive model to obtain the predicted pattern of the obtained financial data after inverse transformation and decoding operations.

Performing inverse transformation and decoding operations on the vector autoregressive model, in an exemplary embodiment of the present application, includes: a topological space S covered by a family of open sets Uα, namely

There is a homeomorphic mapping for any open set Uα:

It is an n-dimensional European space.

Is called coordinate mapping, its inverse mapping

The local coordinate representation called the manifold, in the machine learning framework,

Is called code mapping,

It is called decoding mapping, that is, inverse transformation and decoding operation process.

In this step, when the forecast pattern of the 5th day is needed, the financial data obtained from the 1st to the 4th day is input into the VAR model as the sample data, and the inverse transformation operation may be to compare the obtained forecast data The corresponding sample data is obtained through the mapping and the matrix. The K-line graph related data corresponding to the sample data is the predicted financial data. The decoding operation is the analytical calculation process of the mapping and the matrix. In a feasible implementation scenario of the exemplary embodiment, by inputting the financial data of the previous 4 days into the VAR model, the predicted financial data on the 5th day is inversely transformed and decoded to obtain the local coordinate data, which is compared with the financial data of the previous 4 days. The data is combined to form a predictive candlestick chart.

In this embodiment, based on the law of manifold distribution, a homeomorphic mapping from the image sample space to the feature space is established for the K-line image samples of financial data, and a vector autoregressive model is constructed. When the obtained financial data is input into the model, the out-of-sample After the financial data, the image display of future financial data can be obtained through the calculation of the model, and the financial data can be predicted from the perspective of image recognition. It replaces the traditional method of predicting the rate of return and can be extended to all other types. K-line chart data has a wide range of applications.

2 is a schematic structural diagram of an image processing device for financial data provided by an embodiment of the application. The device can be implemented by software and/or hardware, generally integrated in a smart terminal, and implemented by an image processing method for financial data. As shown in the figure, this embodiment can provide an image processing device for financial data based on the foregoing embodiment, which mainly includes an acquisition module 210, a processing module 220, and a regression module 230.

The obtaining module 210 is configured to obtain the first data Kd at the daily level and the second data Kmin at the preset minute level, where the first data Kd is a multidimensional data type and has a higher dimension than the second data Kmin;

The processing module 220 is configured to perform data processing on the first data Kd of the daily level and the second data Kmin of the preset minute level, and establish a vector autoregressive model of the image sample space according to the processed data;

The regression module 230 is used to input the financial data to be processed into the vector autoregressive model to obtain the predicted pattern of the financial data to be processed after inverse transformation and decoding operations.

In an implementation manner of the exemplary embodiment of the present application, the data type of the first data Kd of the acquisition module includes open price (daily opening price), close price (daily closing data), high price (daily highest) Price) and low price (daily lowest price) financial data in four dimensions, and the second data Kmin includes data in two dimensions: open price and close price.

In an implementation of the exemplary embodiment of the present application, the processing module 220 includes:

The construction sub-module is used to construct the vector autoregressive model:

Construct a homeomorphic mapping of image information of the first data Kd: Formula 1:

This mapping transforms high-dimensional manifolds into low-dimensional manifolds;

Form a local coordinate card;

Dimensional vector (Pr, M, S) time _{_{series: x i = (Pr i,}} M i, S i), Pr i, M i, S i K are the physical size of the i-th line in FIG trading day, Median and intraday amplitude;

The regression equation is:

x _n is the K-line graph on the nth day, c is the three-dimensional constant vector, ε is the error vector, and _Ai is the 3*3 dimensional regression coefficient matrix;

Said inputting the financial data to be processed into the vector autoregressive model after inverse transformation and decoding operation to obtain the prediction pattern of the financial data to be processed includes:

A _i and c are obtained by the regression equation.

The image processing device for financial data provided in the foregoing embodiment can execute the image processing method for financial data provided in any embodiment of this application, and has the corresponding functional modules and beneficial effects for executing the method, which are not detailed in the foregoing embodiment. For the described technical details, please refer to the image processing method for financial data provided in any embodiment of this application.

This embodiment also provides a computer device, such as a smart phone, a tablet computer, a notebook computer, a desktop computer, a rack server, a blade server, a tower server, or a cabinet server (including independent servers, or A server cluster composed of multiple servers), etc. The computer device 20 in this embodiment at least includes but is not limited to: a memory 21 and a processor 22 that can be communicatively connected to each other through a system bus, as shown in FIG. 3. It should be pointed out that FIG. 3 only shows the computer device 20 with components 21-22, but it should be understood that it is not required to implement all the illustrated components, and more or fewer components may be implemented instead.

In this embodiment, the memory 21 (ie, readable storage medium) includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), Read only memory (ROM), electrically erasable programmable read only memory (EEPROM), programmable read only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 21 may be an internal storage unit of the computer device 20, such as a hard disk or memory of the computer device 20. In other embodiments, the memory 21 may also be an external storage device of the computer device 20, such as a plug-in hard disk equipped on the computer device 20, a smart memory card (Smart MediaCard, SMC), and a secure digital (Secure Digital, SD ) Card, Flash Card, etc. Of course, the memory 21 may also include both an internal storage unit of the computer device 20 and an external storage device thereof. In this embodiment, the memory 21 is generally used to store the operating system and various application software installed in the computer device 20, such as the program code of the RNNs neural network in the first embodiment, etc. In addition, the memory 21 can also be used to temporarily store various types of data that have been output or will be output.

The processor 22 may be a central processing unit (Central Processing Unit, CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor 22 is generally used to control the overall operation of the computer device 20. In this embodiment, the processor 22 is used to run the program code or processing data stored in the memory 21, for example, the program code for implementing the image processing method of financial data. When the program code of the image processing method for financial data is executed, the following step:

Acquiring first data Kd at a daily level and second data Kmin at a preset minute level, where the first data Kd is a multidimensional data type and has a higher dimension than the second data Kmin;

This embodiment also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium. Such as flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read only memory (ROM), electrically erasable and programmable memory Read memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disks, optical disks, servers, App application malls, etc., have computer programs stored thereon, and when the computer programs are executed by the processor, the following method steps are implemented :

Another embodiment involving a computer program product includes computer-executable instructions for each of the devices corresponding to at least one of the illustrated systems and/or products. These instructions can be subdivided into subroutines and/or stored in one or more files that may be statically or dynamically linked.

The carrier of the computer program may be any entity or device capable of carrying the program. For example, the carrier may contain a storage medium such as (ROM such as CDROM or semiconductor ROM) or magnetic recording medium (such as floppy disk or hard disk). Further, the carrier may be a transmissible carrier, such as an electrical or optical signal, which may be transmitted via a cable or an optical cable, or by radio or other means. When the program is embodied as such a signal, the carrier may be composed of such a cable or device. Alternatively, the carrier may be an integrated circuit in which the program is embedded, and the integrated circuit is suitable for performing the related method or used for the execution of the related method.

It should be noted that the above-mentioned embodiments illustrate the application, rather than limit the application, and those skilled in the art will be able to design many alternative embodiments without departing from the scope of the appended claims . In the claims, any reference signs placed between parentheses shall not be construed as limiting the claims. The use of the verb "to include" and its conjugations does not exclude the presence of elements or steps other than those recited in the claims. The article "a" or "an" before an element does not exclude the existence of a plurality of such elements. The application can be implemented by hardware including several distinct components, as well as by a suitably programmed computer. In the device claims enumerating several devices, several of these devices can be embodied by the same item of hardware. The mere fact that certain measures are stated in mutually different dependent claims does not indicate that a combination of these measures cannot be used to benefit.

If desired, the different functions discussed herein may be performed in a different order and/or simultaneously with each other. Furthermore, if desired, one or more of the functions described above may be optional or may be combined.

If desired, the steps discussed above are not limited to the execution order in each embodiment, and different steps may be executed in a different order and/or executed simultaneously with each other. In addition, in other embodiments, one or more of the steps described above may be optional or may be combined.

Although various aspects of the application are given in the independent claims, other aspects of the application include combinations of features from the described embodiments and/or dependent claims with features of the independent claims, rather than just the claims. The combination clearly given in.

It should be noted here that although the example embodiments of the present application have been described above, these descriptions should not be understood in a limiting sense. On the contrary, several changes and modifications may be made without departing from the scope of the application as defined in the appended claims.

Those of ordinary skill in the art should understand that each module in the device of the embodiment of the application can be implemented by a general computing device, and each module can be concentrated in a single computing device or a network group composed of computing devices. The device corresponds to the method in the foregoing embodiment, which can be implemented by executable program code, or by a combination of integrated circuits, so this application is not limited to specific hardware or software and a combination thereof.

Those of ordinary skill in the art should understand that each module in the device in the embodiment of this application can be implemented by a universal mobile terminal, and each module can be concentrated in a single mobile terminal or a combination of devices composed of mobile terminals. The device corresponds to the method in the foregoing embodiment, which can be implemented by editing executable program codes, or can be implemented by a combination of integrated circuits, so this application is not limited to specific hardware or software and combinations thereof.

Note that the above are only exemplary embodiments of this application and the technical principles applied. Those skilled in the art will understand that the present application is not limited to the specific embodiments described herein, and various obvious changes, readjustments and substitutions can be made to those skilled in the art without departing from the protection scope of the present application. Therefore, although the application has been described in more detail through the above embodiments, the application is not limited to the above embodiments, and can also include more other equivalent embodiments without departing from the concept of the application. The scope of is determined by the scope of the appended claims.

Claims

An image processing method for financial data, wherein the method includes:

Acquiring first data Kd at a daily level and second data Kmin at a preset minute level, where the first data Kd is a multidimensional data type and has a higher dimension than the second data Kmin;

Data processing is performed on the first data Kd of the daily level and the second data Kmin of the preset minute level, and a vector autoregressive model of the image sample space is established according to the processed data;

The financial data to be processed is input into the vector autoregressive model to obtain a prediction pattern of the financial data to be processed after inverse transformation and decoding operations.
The method according to claim 1, wherein the first data Kd at the daily level and the second data Kmin at the preset minute level are obtained, wherein the first data Kd is a multi-dimensional data type and has a higher dimension than the second data Kmin, including:

The data type of the first data Kd includes four dimensions of financial data of daily opening price, daily closing data, daily highest price, and daily lowest price, and the second data Kmin includes data of two dimensions: open price and close price.
The method according to claim 2, wherein the data processing is performed on the first data Kd of the daily level and the second data Kmin of the preset minute level, and a vector of the image sample space is established according to the processed data Autoregressive models, including:

Constructing the vector autoregressive model includes:

Construct a homeomorphic mapping of the image information of the first data Kd: Formula 1:

This mapping transforms a high-dimensional manifold into a low-dimensional manifold;
Represents the mapping sequence of coordinate mapping for a K-line chart sample with a period of 5 days;
Form a local coordinate card;

Dimensional vector (Pr, M, S) time series: x i = (Pr i, M i, S i), Pr i, M i, S i K are the physical size of the i-th line in FIG trading day, Median and intraday amplitude;

The regression equation is:

x n is the K-line graph on the nth day, c is the three-dimensional constant vector, ε is the error vector, and Ai is the 3*3 dimensional regression coefficient matrix;

Said inputting the financial data to be processed into the vector autoregressive model after inverse transformation and decoding operation to obtain the prediction pattern of the financial data to be processed includes:

A i and c are obtained by the regression equation.
The method according to claim 3, wherein said regression equation derived by A i and c, comprising:

The optimal solution of the regression equation is obtained by the following formula 2:

x i is the sample point, which is an n*m-dimensional matrix in this scheme. The L2 norm of the matrix is defined as the square root of the maximum eigenvalue of x i T x i
Get mapped
Make

R d is a d-dimensional vector space, and {x i } is a set of sample points.
The method according to claim 4, wherein when the optimal solution of the regression equation is obtained by formula 2, the loss function
An image processing device for financial data, wherein the device includes:

The obtaining module is used to obtain the first data Kd at the daily level and the second data Kmin at the preset minute level, wherein the first data Kd is a multidimensional data type and has a higher dimension than the second data Kmin;

A processing module, configured to perform data processing on the first data Kd of the daily level and the second data Kmin of the preset minute level, and establish a vector autoregressive model of the image sample space according to the processed data;

The regression module is used to input the financial data to be processed into the vector autoregressive model to obtain the predicted pattern of the financial data to be processed after inverse transformation and decoding operations.
The device according to claim 6, wherein the data type of the first data Kd of the acquiring module includes financial data in four dimensions: daily opening price, daily closing data, daily highest price, and daily lowest price. The second data Kmin includes data in two dimensions: open price and close price.
The apparatus according to claim 7, wherein the processing module comprises:

The construction sub-module is used to construct the vector autoregressive model:

Construct a homeomorphic mapping of the image information of the first data Kd: Formula 1:

This mapping transforms a high-dimensional manifold into a low-dimensional manifold; the
Represents the mapping sequence of coordinate mapping for a K-line chart sample with a period of 5 days;
Form a local coordinate card;

Dimensional vector (Pr, M, S) time series: x i = (Pr i, M i, S i), Pr i, M i, S i K are the physical size of the i-th line in FIG trading day, Median and intraday amplitude;

The regression equation is:

x n is the K-line graph on the nth day, c is the three-dimensional constant vector, ε is the error vector, and Ai is the 3*3 dimensional regression coefficient matrix;

Regression module, specifically used for:

A i and c are obtained by the regression equation.
The device according to claim 8, wherein the regression module is specifically configured to: when implementing the step of obtaining Ai and c through the regression equation:

The optimal solution of the regression equation is obtained by the following formula 2:

x i is the sample point, which is an n*m-dimensional matrix in this scheme. The L2 norm of the matrix is defined as the square root of the maximum eigenvalue of x i T x i
Get mapped
Make

R d is a d-dimensional vector space, and {x i } is a set of sample points.
The device according to claim 9, wherein when the optimal solution of the regression equation is obtained by formula 2, the loss function
A computer device includes a memory, a processor, and a computer program stored on the memory and running on the processor, wherein the processor implements the following steps when the processor executes the computer program:

Acquiring first data Kd at a daily level and second data Kmin at a preset minute level, where the first data Kd is a multidimensional data type and has a higher dimension than the second data Kmin;

Data processing is performed on the first data Kd of the daily level and the second data Kmin of the preset minute level, and a vector autoregressive model of the image sample space is established according to the processed data;

The financial data to be processed is input into the vector autoregressive model to obtain a prediction pattern of the financial data to be processed after inverse transformation and decoding operations.
The computer device according to claim 11, wherein the first data Kd at the daily level and the second data Kmin at the preset minute level are acquired, wherein the first data Kd is a multidimensional data type and has a higher dimension than The steps of the second data Kmin specifically include:

The data type of the first data Kd includes four dimensions of financial data of daily opening price, daily closing data, daily highest price, and daily lowest price, and the second data Kmin includes data of two dimensions: open price and close price.
The computer device according to claim 12, wherein the first data Kd of the daily level and the second data Kmin of the preset minute level are processed after the realization of the data processing, and an image sample is created based on the processed data The steps of the spatial vector autoregressive model include:

Constructing the vector autoregressive model includes:

Construct a homeomorphic mapping of the image information of the first data Kd: Formula 1:

This mapping transforms a high-dimensional manifold into a low-dimensional manifold; the
Represents the mapping sequence of coordinate mapping for a K-line chart sample with a period of 5 days;
Form a local coordinate card;

Dimensional vector (Pr, M, S) time series: x i = (Pr i, M i, S i), Pr i, M i, S i K are the physical size of the i-th line in FIG trading day, Median and intraday amplitude;

The regression equation is:

x n is the K-line graph on the nth day, c is the three-dimensional constant vector, ε is the error vector, and Ai is the 3*3 dimensional regression coefficient matrix;

When implementing the step of inputting the financial data to be processed into the vector autoregressive model after inverse transformation and decoding operations to obtain the predicted pattern of the financial data to be processed, it specifically includes:

A i and c are obtained by the regression equation.
The computer device according to claim 13, wherein when the step of obtaining A i and c by the regression equation is implemented, it specifically comprises:

The optimal solution of the regression equation is obtained by the following formula 2:

x i is the sample point, which is an n*m-dimensional matrix in this scheme. The L2 norm of the matrix is defined as the square root of the maximum eigenvalue of x i T x i
Get mapped
Make

R d is a d-dimensional vector space, and {x i } is a set of sample points.
The computer device according to claim 14, wherein when the optimal solution of the regression equation is obtained by formula 2, the loss function
A computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the following steps when executed by a processor:

Acquiring first data Kd at a daily level and second data Kmin at a preset minute level, where the first data Kd is a multidimensional data type and has a higher dimension than the second data Kmin;

Data processing is performed on the first data Kd of the daily level and the second data Kmin of the preset minute level, and a vector autoregressive model of the image sample space is established according to the processed data;

The financial data to be processed is input into the vector autoregressive model to obtain a prediction pattern of the financial data to be processed after inverse transformation and decoding operations.
The computer-readable storage medium according to claim 16, wherein the first data Kd at the daily level and the second data Kmin at the preset minute level are acquired in the realization of the acquisition, wherein the first data Kd is a multi-dimensional data type and When the dimension is higher than the second data Kmin, the steps specifically include:

The data type of the first data Kd includes four dimensions of financial data of daily opening price, daily closing data, daily highest price, and daily lowest price, and the second data Kmin includes data of two dimensions: open price and close price.
The computer-readable storage medium according to claim 17, wherein the first data Kd of the daily level and the second data Kmin of the preset minute level are processed in the realization of the data processing, and the data is processed according to the processed data. The steps of establishing the vector autoregressive model of the image sample space include:

Constructing the vector autoregressive model includes:

Construct a homeomorphic mapping of the image information of the first data Kd: Formula 1:

This mapping transforms a high-dimensional manifold into a low-dimensional manifold; the
Represents the mapping sequence of coordinate mapping for a K-line chart sample with a period of 5 days;
Form a local coordinate card;

Dimensional vector (Pr, M, S) time series: x i = (Pr i, M i, S i), Pr i, M i, S i K are the physical size of the i-th line in FIG trading day, Median and intraday amplitude;

The regression equation is:

x n is the K-line graph on the nth day, c is the three-dimensional constant vector, ε is the error vector, and Ai is the 3*3 dimensional regression coefficient matrix;

When implementing the step of inputting the financial data to be processed into the vector autoregressive model after inverse transformation and decoding operations to obtain the predicted pattern of the financial data to be processed, it specifically includes:

A i and c are obtained by the regression equation.
18. The computer-readable storage medium according to claim 18, wherein when the step of obtaining Ai and c from the regression equation is implemented, it specifically comprises:

The optimal solution of the regression equation is obtained by the following formula 2:

x i is the sample point, which is an n*m-dimensional matrix in this scheme. The L2 norm of the matrix is defined as the square root of the maximum eigenvalue of x i T x i
Get mapped
Make

R d is a d-dimensional vector space, and {x i } is a set of sample points.
The computer-readable storage medium according to claim 19, wherein when the optimal solution of the regression equation is obtained by formula 2, the loss function