High-frequency data processing method and device and high-frequency data application method and device

Abstract

The invention relates to the technical field of data processing, and discloses a high-frequency data processing and application method and device. According to the method, a target high-frequency data source is accessed; when the high-frequency data is changed, acquiring the Level2 data through the subscribed interface; performing data synthesis on the Level2 data to obtain a plurality of market information data; performing data verification on each piece of market data; if the data verification is not passed, performing data completion on the market information data by adopting third-party data; if the data verification is passed, determining the market data as high-frequency basic data; compiling a plurality of factor logics by utilizing a programming language according to preset factor business logics; calculating a plurality of factor values based on the high-frequency basic data and a plurality of factor logics; and optimizing the quantitative multi-factor prediction model by using the plurality of factor values. According to the method, high-frequency data cleaning is realized, the quality of the high-frequency data is improved, the programming language limitation is broken through, and the optimization effect of the quantitative multi-factor prediction model is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and device for processing and applying high-frequency data. Background Art

[0002] At present, the quantitative modeling effect based on low-frequency quantity, price and fundamentals is relatively low. High-frequency factor mining requires a high-performance data base and data processing efficiency to support it. The original research model based on Python programming and relational databases cannot store, process and mine high-frequency factor data.

[0003] Regarding the processing of high-frequency data, traditional databases have serious performance bottlenecks, and the amount of Level 2 data far exceeds the limit of traditional relational database data set processing. High-frequency factor mining requires high computing power, and the traditional model of separating "storage" and "computation" is no longer applicable. Traditional quantitative researchers mostly use python to process data locally, but high-frequency factor mining requires massive amounts of data, and researchers simply cannot process large-scale data locally and perform complex concurrent calculations and I / O operations. Intraday high-frequency quantitative investment strategies have very high requirements for the timeliness performance of factor mining, and the existing technical architecture is difficult to support high-frequency investment strategies represented by T+0. Summary of the invention

[0004] The present invention provides a method and device for processing and applying high-frequency data, which realizes high-frequency data cleaning, improves the quality of high-frequency data, breaks the limitation of programming language, and improves the optimization effect of quantitative multi-factor prediction model.

[0005] In order to solve the above technical problems, a first aspect of the present invention provides a method for processing high-frequency data, comprising:

[0006] Access to target high-frequency data sources;

[0007] When high-frequency data changes, obtain Level2 data through the subscribed interface;

[0008] Performing data synthesis on the Level 2 data to obtain several types of market data;

[0009] Performing data verification on each of the market data;

[0010] If the data verification fails, the market data is supplemented with third-party data;

[0011] If the data verification passes, the market data is determined as high-frequency basic data.

[0012] Furthermore, when the high-frequency data changes, the Level 2 data is obtained through the subscribed interface, specifically:

[0013] When an MQ message is received, parsing the MQ message;

[0014] When the data change notification is obtained by parsing the MQ message, the Tick data, K-line data, transaction data, entrustment data and entrustment queue data are obtained through the subscribed interface respectively;

[0015] The tick data, the K-line data, the transaction data, the entrustment data and the entrustment queue data are determined as Level 2 data.

[0016] Furthermore, the Level 2 data is synthesized to obtain several market data, specifically:

[0017] Completing the elements of the K-line data based on the transaction data to obtain updated K-line data;

[0018] The transaction amount and volume for the day are synthesized based on the Tick data, the transaction data, the entrustment data, the entrustment queue data and the updated K-line data.

[0019] Furthermore, the data verification of each of the market data is specifically performed as follows:

[0020] Obtain the basic market data of the day from third-party data sources;

[0021] Compare each of the above market data with the basic market data of the day;

[0022] When the comparison result shows that there is a difference, it is determined that the data verification fails.

[0023] Furthermore, the basic market data includes the opening price of the day, the highest price of the day, the lowest price of the day, the closing price of the day and the trading volume of the day.

[0024] Furthermore, if the data verification fails, the market data is supplemented with third-party data, specifically:

[0025] If the data verification fails, the market data that differs from the basic market data of the day will be determined as data to be replaced;

[0026] Acquire replacement market data corresponding to the data to be replaced from a third-party data source;

[0027] The data to be replaced is replaced with the replacement market data.

[0028] The first aspect of the present invention provides a method for processing high-frequency data, which includes accessing a target high-frequency data source; obtaining Level 2 data through a subscribed interface when the high-frequency data changes; synthesizing the Level 2 data to obtain several types of market data; performing data verification on each market data; if the data verification fails, using third-party data to complete the market data; if the data verification passes, determining the market data as high-frequency basic data. The present invention realizes high-frequency data cleaning and improves the accuracy and quality of high-frequency data.

[0029] A second aspect of the present invention provides a method for applying high-frequency data, comprising:

[0030] According to the high-frequency data processing method of the first aspect of the present invention, high-frequency basic data is obtained;

[0031] According to the preset factor business logic, use programming language to write several factor logics;

[0032] Calculating several factor values ​​based on the high-frequency basic data and several factor logics;

[0033] The quantitative multi-factor prediction model is optimized using several of the factor values.

[0034] Furthermore, based on the high-frequency basic data and the several factor logics, several factor values ​​are calculated, specifically:

[0035] logically transferring the several factors to a time series database;

[0036] Several factor values ​​are calculated in the time series database based on the high-frequency basic data.

[0037] The second aspect of the present invention provides an application method for high-frequency data, which obtains high-frequency basic data; uses a programming language to write several kinds of factor logic according to preset factor business logic; calculates several kinds of factor values ​​based on high-frequency basic data and several kinds of factor logic; and optimizes the quantitative multi-factor prediction model using several kinds of factor values. The present invention uses high-frequency data for mining, improves the optimization effect of the model, breaks the limitation of programming languages, supports multi-paradigm programming languages, and lowers the threshold for business factor development.

[0038] A third aspect of the present invention provides a high-frequency data processing device, comprising: a data source access module, a data acquisition module, a data synthesis module, a data verification module, a data completion module and a data determination module;

[0039] The connection module is used for the data source access module, the data acquisition module, the data synthesis module, the data verification module, the data completion module and the data determination module;

[0040] The data source access module is used to access the target high-frequency data source;

[0041] The data acquisition module is used to obtain Level 2 data through the subscribed interface when the high-frequency data changes;

[0042] The data synthesis module is used to synthesize the Level 2 data to obtain several types of market data;

[0043] The data verification module is used to perform data verification on each of the market data;

[0044] The data completion module is used to complete the market data using third-party data if the data verification fails;

[0045] The data determination module is used to determine the market data as high-frequency basic data if the data verification passes.

[0046] The third aspect of the present invention provides a high-frequency data processing device, which is based on the organic combination of modules and accesses the target high-frequency data source; when the high-frequency data changes, the Level2 data is obtained through the subscribed interface; the Level2 data is synthesized to obtain several types of market data; each market data is verified; if the data verification fails, the market data is supplemented with third-party data; if the data verification passes, the market data is determined as high-frequency basic data. The present invention realizes high-frequency data cleaning and improves the accuracy and quality of high-frequency data.

[0047] A fourth aspect of the present invention provides a high-frequency data processing device, comprising: a basic data acquisition module, a factor logic editing module, a factor value calculation module and a model optimization module;

[0048] The basic data acquisition module is used to acquire high-frequency basic data according to the high-frequency data processing method according to the first aspect of the present invention;

[0049] The factor logic editing module is used to write several kinds of factor logics using a programming language according to the preset factor business logic;

[0050] The factor value calculation module is used to calculate several factor values ​​based on the high-frequency basic data and several factor logics;

[0051] The model optimization module is used to optimize the quantitative multi-factor prediction model using several factor values.

[0052] The fourth aspect of the present invention provides a high-frequency data processing device, which obtains high-frequency basic data based on the organic combination of modules; uses a programming language to write several factor logics according to the preset factor business logic; calculates several factor values ​​based on the high-frequency basic data and several factor logics; and optimizes the quantitative multi-factor prediction model using several factor values. The present invention uses high-frequency data for mining, improves the optimization effect of the model; breaks the limitation of programming languages, supports multi-paradigm programming languages, and lowers the threshold for business factor development. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A schematic diagram of a flow chart of an embodiment of a method for processing high-frequency data provided by the present invention;

[0054] Figure 2 A schematic diagram of a flow chart of an embodiment of a high-frequency data application method provided by the present invention;

[0055] Figure 3 A schematic flow chart of an embodiment of a method for processing and applying high-frequency data provided by the present invention;

[0056] Figure 4 A schematic diagram of the structure of an embodiment of a high-frequency data processing device provided by the present invention;

[0057] Figure 5 A schematic structural diagram of an embodiment of a high-frequency data application device provided by the present invention. DETAILED DESCRIPTION

[0058] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0059] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0060] Some embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0061] See also Figure 1, is a flow chart of an embodiment of a method for processing high-frequency data provided by the present invention, the method includes steps 101 to 106, each step is specifically as follows:

[0062] Step 101: Access a target high-frequency data source.

[0063] In an embodiment of the present invention, the data in the target high-frequency data source will be updated at a fixed frequency. After accessing the target high-frequency data source, message notifications and change data of the target high-frequency data source can be obtained in real time, which can improve the real-time nature of the data and strengthen the connection with the target high-frequency data source.

[0064] Step 102: When the high-frequency data changes, obtain the Level 2 data through the subscribed interface.

[0065] Further, in the embodiment of the present invention, when the high-frequency data changes, the Level 2 data is obtained through the subscribed interface, specifically:

[0066] When an MQ message is received, parsing the MQ message;

[0067] When the data change notification is obtained by parsing the MQ message, the Tick data, K-line data, transaction data, entrustment data and entrustment queue data are obtained through the subscribed interface respectively;

[0068] The tick data, the K-line data, the transaction data, the entrustment data and the entrustment queue data are determined as Level 2 data.

[0069] In an embodiment of the present invention, after accessing the target high-frequency data source, the target high-frequency data source can interact through MQ messages. When an MQ message sent by the target high-frequency data source is received, the notification transmitted by the target high-frequency data source can be learned by parsing the MQ message. When the data change notification is parsed, the corresponding data can be obtained through the subscribed interface. Among them, the subscribed interface is a way to obtain the corresponding data of the target high-frequency data source according to demand. Specifically, the subscribed interface includes market snapshots, L2 transaction by transaction, L2 entrustment by transaction, etc. Through the subscribed interface, Tick data, K-line data, transaction by transaction data, entrustment data and entrustment queue data, etc. can be obtained from the target high-frequency data source.

[0070] Step 103: synthesize the Level 2 data to obtain several types of market data.

[0071] Furthermore, in the embodiment of the present invention, the Level 2 data is synthesized to obtain several types of market data, specifically:

[0072] Completing the elements of the K-line data based on the transaction data to obtain updated K-line data;

[0073] The transaction amount and volume for the day are synthesized based on the Tick data, the transaction data, the entrustment data, the entrustment queue data and the updated K-line data.

[0074] In the embodiment of the present invention, the K-line data obtained through the subscribed interface does not include the number of transactions, so the transaction number data of the K-line can be generated by the transaction data obtained through the subscribed interface, and the K-line data containing the transaction number can be updated by combining the transaction number data of the K-line to complete data alignment and element completion. After the obtained data is preliminarily processed, data synthesis can be performed to obtain the transaction amount and volume of the day.

[0075] Step 104: Perform data verification on each of the market data.

[0076] Furthermore, in the embodiment of the present invention, data verification is performed on each of the market data, specifically:

[0077] Obtain the basic market data of the day from third-party data sources;

[0078] Compare each of the above market data with the basic market data of the day;

[0079] When the comparison result shows that there is a difference, it is determined that the data verification fails.

[0080] Furthermore, in an embodiment of the present invention, the basic market data includes the opening price of the day, the highest price of the day, the lowest price of the day, the closing price of the day and the trading volume of the day.

[0081] In the embodiment of the present invention, after obtaining the data of the target high-frequency data source through the subscribed interface and performing preliminary processing on the obtained data, the obtained data is verified according to the third-party data to ensure the accuracy of the data. Specifically, the third-party data may include the opening price of the day, the highest price of the day, the lowest price of the day, the closing price of the day and the trading volume of the day.

[0082] Step 105: If the data verification fails, third-party data is used to complete the market data.

[0083] Further, in the embodiment of the present invention, if the data verification fails, the market data is supplemented with third-party data, specifically:

[0084] If the data verification fails, the market data that differs from the basic market data of the day will be determined as data to be replaced;

[0085] Acquire replacement market data corresponding to the data to be replaced from a third-party data source;

[0086] The data to be replaced is replaced with the replacement market data.

[0087] In an embodiment of the present invention, during the data verification process, if it is detected that the acquired market data differs from the third-party data, or some data is missing, the full market data for the day is corrected and supplemented based on the third-party data, thereby improving the accuracy of the high-frequency data.

[0088] Step 106: If the data verification passes, the market data is determined as high-frequency basic data.

[0089] In the embodiment of the present invention, after the high-frequency data obtained from the target high-frequency data source completes data verification, the market data of the convertible bond is cleaned, and the high-frequency data after data verification can become the basic data for factor mining. Among them, the basic data needs to be cleaned daily, so the operations of obtaining high-frequency data, performing preliminary processing on the high-frequency data, and performing data verification on the high-frequency data are repeated every day to ensure the real-time and correctness of the high-frequency data.

[0090] In summary, the embodiment of the present invention provides a method for processing high-frequency data, accessing a target high-frequency data source; when high-frequency data changes, obtaining Level2 data through a subscribed interface; synthesizing Level2 data to obtain several types of market data; performing data verification on each market data; if the data verification fails, using third-party data to complete the market data; if the data verification passes, determining the market data as high-frequency basic data. The present invention realizes high-frequency data cleaning and improves the accuracy and quality of high-frequency data.

[0091] See also Figure 2 , is a flow chart of an embodiment of a high-frequency data application method provided by the present invention, the method includes steps 201 to 204, each step is specifically as follows:

[0092] Step 201: According to the high-frequency data processing method described in the first aspect of the present invention, high-frequency basic data is obtained.

[0093] In the embodiment of the present invention, high-frequency basic data is obtained by using the high-frequency data processing method described in the first aspect of the present invention, which can ensure the quality of the high-frequency data and improve the efficiency and accuracy of factor mining.

[0094] Step 202: According to the preset factor business logic, use a programming language to write several types of factor logic.

[0095] In the embodiment of the present invention, the present invention supports multi-paradigm programming languages. Business personnel can write factors in python or SQL or other programming languages ​​according to the factor business logic, derive multiple factor logics, and complete factor editing. For example, the calculation formula for calculating the factor of the proportion of transactions at the end of the day based on step-by-step transactions can be: the factor of the proportion of transactions at the end of the day = the transaction volume at the end of the half hour / the total transaction volume of the day.

[0096] Step 203: Calculate several factor values ​​based on the high-frequency basic data and several factor logics.

[0097] Furthermore, in the embodiment of the present invention, based on the high-frequency basic data and the several factor logics, several factor values ​​are calculated, specifically:

[0098] logically transferring the several factors to a time series database;

[0099] Several factor values ​​are calculated in the time series database based on the high-frequency basic data.

[0100] In the embodiment of the present invention, after obtaining the written factor logic, the factor logic is transmitted to the time series database computing node by calling the time series database computing engine interface. After receiving the call, the time series database obtains multiple factor values ​​by parsing and executing the factor logic.

[0101] Step 204: Optimize the quantitative multi-factor prediction model using several of the factor values.

[0102] In an embodiment of the present invention, the factor value obtained based on the high-frequency data is introduced into the quantitative multi-factor prediction model, and the factor Alpha value can be calculated, thereby optimizing the quantitative multi-factor prediction model and improving the quantitative modeling effect.

[0103] In summary, the embodiment of the present invention provides an application method for high-frequency data, which obtains high-frequency basic data; uses a programming language to write several types of factor logic according to the preset factor business logic; calculates several types of factor values ​​based on the high-frequency basic data and several types of factor logic; and uses several types of factor values ​​to optimize the quantitative multi-factor prediction model. The present invention uses high-frequency data for mining, which improves the optimization effect of the model; breaks the limitation of programming languages, supports multi-paradigm programming languages, and lowers the threshold for business factor development.

[0104] As an example of an embodiment of the present invention, see Figure 3, is a flow chart of an embodiment of the high-frequency data processing and application method provided by the present invention. Access a high-frequency data source, complete the market interface subscription in the high-frequency data source; obtain and parse market information in real time through the subscribed interface; align and complete the acquired market information to complete the preliminary processing of the market data; then perform data verification on the market data after preliminary processing. If the data verification fails, use a third-party data source to supplement the data until the data verification passes, forming high-frequency basic data and completing high-frequency data cleaning. The high-frequency basic data obtained according to the above process can be used for factor mining. After factor editing and factor calculation based on the high-frequency basic data, the factor value can be obtained, the factor Alpha value is calculated based on the factor value, and the quantitative multi-factor prediction model is optimized to improve the operation effect of the model.

[0105] See also Figure 4 , is a schematic diagram of the structure of an embodiment of a high-frequency data processing device provided by the present invention, the device includes a data source access module 301, a data acquisition module 302, a data synthesis module 303, a data verification module 304, a data completion module 305 and a data determination module 306;

[0106] The data source access module 301 is used to access the target high-frequency data source;

[0107] The data acquisition module 302 is used to obtain Level 2 data through the subscribed interface when the high-frequency data changes;

[0108] The data synthesis module 303 is used to synthesize the Level 2 data to obtain several market data;

[0109] The data verification module 304 is used to perform data verification on each of the market data;

[0110] The data completion module 305 is used to complete the market data using third-party data if the data verification fails;

[0111] The data determination module 306 is used to determine the market data as high-frequency basic data if the data verification passes.

[0112] Further, in the embodiment of the present invention, when the high-frequency data changes, the Level 2 data is obtained through the subscribed interface, specifically:

[0113] When an MQ message is received, parsing the MQ message;

[0114] When the data change notification is obtained by parsing the MQ message, the Tick data, K-line data, transaction data, entrustment data and entrustment queue data are obtained through the subscribed interface respectively;

[0115] The tick data, the K-line data, the transaction data, the entrustment data and the entrustment queue data are determined as Level 2 data.

[0116] Furthermore, in the embodiment of the present invention, the Level 2 data is synthesized to obtain several types of market data, specifically:

[0117] Completing the elements of the K-line data based on the transaction data to obtain updated K-line data;

[0118] The transaction amount and volume for the day are synthesized based on the Tick data, the transaction data, the entrustment data, the entrustment queue data and the updated K-line data.

[0119] Furthermore, in the embodiment of the present invention, data verification is performed on each of the market data, specifically:

[0120] Obtain the basic market data of the day from third-party data sources;

[0121] Compare each of the above market data with the basic market data of the day;

[0122] When the comparison result shows that there is a difference, it is determined that the data verification fails.

[0123] Furthermore, in an embodiment of the present invention, the basic market data includes the opening price of the day, the highest price of the day, the lowest price of the day, the closing price of the day and the trading volume of the day.

[0124] Further, in the embodiment of the present invention, if the data verification fails, the market data is supplemented with third-party data, specifically:

[0125] If the data verification fails, the market data that differs from the basic market data of the day will be determined as data to be replaced;

[0126] Acquire replacement market data corresponding to the data to be replaced from a third-party data source;

[0127] The data to be replaced is replaced with the replacement market data.

[0128] In summary, the embodiment of the present invention provides a high-frequency data processing device, which is based on the organic combination of modules and accesses the target high-frequency data source; when the high-frequency data changes, the Level2 data is obtained through the subscribed interface; the Level2 data is synthesized to obtain several types of market data; each market data is verified; if the data verification fails, the market data is supplemented with third-party data; if the data verification passes, the market data is determined as high-frequency basic data. The present invention realizes high-frequency data cleaning and improves the accuracy and quality of high-frequency data.

[0129] See also Figure 5 , is a structural diagram of an embodiment of a high-frequency data application device provided by the present invention, the device includes a basic data acquisition module 401, a factor logic editing module 402, a factor value calculation module 403 and a model optimization module 404;

[0130] The basic data acquisition module 401 is used to acquire high-frequency basic data according to the high-frequency data processing method described in the first aspect of the present invention;

[0131] The factor logic editing module 402 is used to write several types of factor logics using a programming language according to the preset factor business logic;

[0132] The factor value calculation module 403 is used to calculate several factor values ​​based on the high-frequency basic data and several factor logics;

[0133] The model optimization module 404 is used to optimize the quantitative multi-factor prediction model using several of the factor values.

[0134] Furthermore, in the embodiment of the present invention, based on the high-frequency basic data and the several factor logics, several factor values ​​are calculated, specifically:

[0135] logically transferring the several factors to a time series database;

[0136] Several factor values ​​are calculated in the time series database based on the high-frequency basic data.

[0137] In summary, the embodiment of the present invention provides an application device for high-frequency data, which obtains high-frequency basic data based on the organic combination of modules; uses a programming language to write several types of factor logic according to the preset factor business logic; calculates several types of factor values ​​based on the high-frequency basic data and several types of factor logic; and optimizes the quantitative multi-factor prediction model using several types of factor values. The present invention uses high-frequency data for mining, improves the optimization effect of the model; breaks the limitation of programming languages, supports multi-paradigm programming languages, and lowers the threshold for business factor development.

[0138] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for processing high-frequency data, characterized in that: include: Access to target high-frequency data sources; When high-frequency data changes, obtain Level2 data through the subscribed interface; Performing data synthesis on the Level 2 data to obtain several types of market data; Performing data verification on each of the market data; If the data verification fails, the market data is supplemented with third-party data; If the data verification passes, the market data is determined as high-frequency basic data.

2. The method for processing high-frequency data according to claim 1, characterized in that: When the high-frequency data changes, the Level 2 data is obtained through the subscribed interface, specifically: When an MQ message is received, parsing the MQ message; When the data change notification is obtained by parsing the MQ message, the Tick data, K-line data, transaction data, entrustment data and entrustment queue data are obtained through the subscribed interface respectively; The tick data, the K-line data, the transaction data, the entrustment data and the entrustment queue data are determined as Level 2 data.

3. The method for processing high-frequency data according to claim 2, characterized in that: The Level 2 data is synthesized to obtain several market data, specifically: Completing the elements of the K-line data based on the transaction data to obtain updated K-line data; The transaction amount and volume for the day are synthesized based on the Tick data, the transaction data, the entrustment data, the entrustment queue data and the updated K-line data.

4. The method for processing high frequency data according to claim 3, characterized in that: The data verification of each of the market data is specifically as follows: Obtain the basic market data of the day from third-party data sources; Compare each of the above market data with the basic market data of the day; When the comparison result shows that there is a difference, it is determined that the data verification fails.

5. The method for processing high frequency data according to claim 4, characterized in that: The basic market data include the opening price of the day, the highest price of the day, the lowest price of the day, the closing price of the day and the trading volume of the day.

6. The method for processing high frequency data according to claim 4, characterized in that: If the data verification fails, the market data is supplemented with third-party data, specifically: If the data verification fails, the market data that differs from the basic market data of the day will be determined as data to be replaced; Acquire replacement market data corresponding to the data to be replaced from a third-party data source; The data to be replaced is replaced with the replacement market data.

7. A method for applying high-frequency data, characterized in that: include: According to the high-frequency data processing method according to any one of claims 1 to 6, obtaining high-frequency basic data; According to the preset factor business logic, use programming language to write several factor logics; Calculating several factor values ​​based on the high-frequency basic data and several factor logics; The quantitative multi-factor prediction model is optimized using several of the factor values.

8. The high frequency data application method according to claim 7, characterized in that: The calculation of several factor values ​​based on the high-frequency basic data and several factor logics is specifically as follows: logically transferring the several factors to a time series database; Several factor values ​​are calculated in the time series database based on the high-frequency basic data.

9. A high-frequency data processing device, characterized in that: include: Data source access module, data acquisition module, data synthesis module, data verification module, data completion module and data determination module; The data source access module is used to access the target high-frequency data source; The data acquisition module is used to obtain Level 2 data through the subscribed interface when the high-frequency data changes; The data synthesis module is used to synthesize the Level 2 data to obtain several types of market data; The data verification module is used to perform data verification on each of the market data; The data completion module is used to complete the market data using third-party data if the data verification fails; The data determination module is used to determine the market data as high-frequency basic data if the data verification passes.

10. An application device for high-frequency data, characterized in that: include: Basic data acquisition module, factor logic editing module, factor value calculation module and model optimization module; The basic data acquisition module is used to acquire high-frequency basic data according to the high-frequency data processing method according to any one of claims 1 to 6; The factor logic editing module is used to write several types of factor logics using a programming language according to the preset factor business logic; The factor value calculation module is used to calculate several factor values ​​based on the high-frequency basic data and several factor logics; The model optimization module is used to optimize the quantitative multi-factor prediction model using several factor values.

Images