WO2020024712A1

WO2020024712A1 - Method and device for predicting number of foreign transactions

Info

Publication number: WO2020024712A1
Application number: PCT/CN2019/091570
Authority: WO
Inventors: 杨永晟
Original assignee: 阿里巴巴集团控股有限公司
Priority date: 2018-08-02
Filing date: 2019-06-17
Publication date: 2020-02-06
Also published as: TW202008231A; TWI714113B; CN108846525A

Abstract

A method and device for predicting the number of foreign transactions, the method comprising: using a plurality of prediction models, predicting the number of foreign transactions for each business day from different perspectives, and then according to a prediction value and an actual value, calculating a variation trend for the number of foreign transactions, thereby predicting the number of foreign transactions within a future time period, and increasing the accuracy of prediction.

Description

Method and device for predicting foreign exchange transaction volume

Technical field

This specification relates to the field of data processing technology, and particularly to a method and device for predicting foreign exchange transaction volume.

Background technique

In the international currency exchange business, it is necessary to reduce the potential risk of fluctuations in exchange rate exposure and control profit and loss by purchasing the foreign exchange transaction volume of the next foreign exchange purchase settlement cycle in advance. In order to carry out profit and loss control, it is necessary to forecast the foreign exchange transaction volume of each foreign exchange settlement cycle. Therefore, it is necessary to improve the forecasting method of foreign exchange transaction volume.

Summary of the invention

Based on this, this manual provides methods and devices for predicting foreign exchange transaction volume.

According to a first aspect of the embodiments of the present specification, a method for predicting foreign exchange transaction volume is provided. The method includes: calculating a foreign exchange transaction volume according to a predicted foreign exchange transaction volume value for each business day and a real foreign exchange transaction volume value corresponding to the business day. Change trends, and predict foreign exchange transaction volume in the future time period according to the change trends, wherein the predicted foreign exchange transaction volume for each business day is obtained according to the following methods: historical foreign exchange transaction volume before the current forecast time node Input multiple pre-trained prediction models to obtain the current first prediction result corresponding to each prediction model; calculate the predicted value of foreign exchange transaction volume on the Nth business day after the current prediction time node according to each current first prediction result; N is A preset positive integer.

Optionally, the step of calculating a foreign exchange transaction volume prediction value on the Nth business day after the current prediction time node according to each current first prediction result includes: using a weighted average value of each current first prediction result as the current prediction. Forecast value of foreign exchange transaction volume on the Nth business day after the time node.

Optionally, the method further comprises: calculating an impact factor of the specific business day according to the predicted value of the foreign exchange transaction volume on a specific business day; and performing a prediction on the predicted value of the foreign exchange transaction volume of the specific business day in a future time period according to the impact factor. Amended.

Optionally, the step of calculating the impact factor of the specific business day based on the predicted value of the foreign exchange transaction volume on the specific business day includes: comparing the predicted value of the foreign exchange transaction volume on the specific business day with the foreign exchange transaction volume on the specific business day. The ratio of the true values is used as the influence factor.

Optionally, the method further includes: separately inputting the foreign exchange transaction volume of a specific business day into the plurality of prediction models to obtain second prediction results corresponding to each prediction model; and calculating the specific business day according to each second prediction result Forex trading volume forecast.

Optionally, the step of calculating a foreign exchange transaction volume prediction value on the specific business day according to each second prediction result includes: using a weighted average of each second prediction result as the foreign exchange transaction volume prediction value on the specific business day.

Optionally, the historical foreign exchange transaction volume before the current prediction time node is respectively input into a plurality of pre-trained prediction models, and the step of obtaining the current first prediction result corresponding to each prediction model includes: entering the historical foreign exchange transaction before the current prediction time node. Input the pre-trained first prediction model to obtain the first prediction result corresponding to the first prediction model; input the historical foreign exchange transaction volume of the N business days closest to the current prediction time node into the pre-trained second prediction model to obtain the second A first prediction result corresponding to the prediction model; and a historical foreign exchange transaction volume of N business days in the historical period of the current prediction time node is input to a pre-trained third prediction model to obtain a first prediction result corresponding to the third prediction model.

Optionally, the first prediction model is an ARIMA model, the second prediction model is a mean estimation model, and the third prediction model is a ring-by-trend trend estimation model.

According to a second aspect of the embodiments of the present specification, a device for predicting foreign exchange transaction volume is provided. The device includes a prediction module for respectively predicting a foreign exchange transaction volume value for each business day and a real foreign exchange transaction volume value for a corresponding business day. Calculate the change trend of foreign exchange transaction volume, and predict the foreign exchange transaction volume in the future time period according to the change trend, wherein the forecast value of foreign exchange transaction volume for each business day is obtained by executing the functions of the following modules: input module, It is used to input the historical foreign exchange transaction volume before the current prediction time node into multiple pre-trained prediction models to obtain the current first prediction result corresponding to each prediction model; a calculation module is configured to calculate the first prediction result according to each current first prediction result The forecast value of the foreign exchange transaction volume on the Nth business day after the current forecast time node; N is a preset positive integer.

According to a third aspect of the embodiments of the present specification, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the method according to any one of the embodiments is implemented.

According to a fourth aspect of the embodiments of the present specification, there is provided a computer device including a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor implements any implementation when the program is executed. Example method.

Applying the solution of the embodiment of this specification, using multiple forecasting models, forecasting the foreign exchange transaction volume of each business day from different perspectives, and then calculating the change trend of foreign exchange transaction volume based on the predicted value and the real value, so as to predict the foreign exchange volume in the future time Prediction of transaction volume improves prediction accuracy.

It should be understood that the above general description and the following detailed description are merely exemplary and explanatory, and should not limit the present specification.

BRIEF DESCRIPTION OF THE DRAWINGS

The drawings herein are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the specification, and together with the description serve to explain the principles of the specification.

FIG. 1 is a flowchart of a method for predicting foreign exchange transaction volume according to an embodiment of the present specification.

FIG. 2 is a schematic diagram of time series data according to an embodiment of the present specification.

FIG. 3 is an overall principle diagram of foreign exchange transaction volume prediction according to an embodiment of the present specification.

FIG. 4 is a block diagram of a foreign exchange transaction volume prediction device according to an embodiment of the present specification.

FIG. 5 is a schematic diagram of a computer device for implementing a method of an embodiment of the present specification according to an embodiment of the present specification.

detailed description

Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with this specification. Rather, they are merely examples of devices and methods consistent with certain aspects of the specification, as detailed in the appended claims.

The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the specification. As used in this specification and the appended claims, the singular forms "a", "the" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in this specification to describe various information, the information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of this specification, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein can be interpreted as "at" or "when" or "in response to determination".

As shown in FIG. 1, it is a flowchart of a method for predicting foreign exchange transaction volume according to an embodiment of the present specification. The method may include:

Step 102: Calculate the change trend of the foreign exchange transaction volume according to the predicted value of the foreign exchange transaction volume of each business day and the real value of the foreign exchange transaction volume of the corresponding business day, and predict the foreign exchange transaction volume in the future period according to the change trend. Among them, the forecast value of foreign exchange transaction volume for each business day is obtained according to the following methods:

Step 104: input the historical foreign exchange transaction volume before the current prediction time node into a plurality of pre-trained prediction models, and obtain the current first prediction result corresponding to each prediction model;

Step 106: Calculate the predicted value of the foreign exchange transaction volume on the Nth business day after the current prediction time node according to each current first prediction result; N is a preset positive integer.

In the above embodiment, the business day is a statistical period of the business volume, which may be the same as the natural day (0:00 to 23:59), or may be preset by the business system. For example, it may be set to a certain natural day. 15:00 to 14:59 of the next day, or set to another time period.

In step 102, the predicted value of the foreign exchange transaction volume and the true value of the foreign exchange transaction volume may be obtained for a plurality of business days. For example, (N1, N2, ..., NT) the predicted value of the foreign exchange transaction volume for a total of T business days and The real value, predicted value and real value of the foreign exchange transaction volume are set to (v1, v2, ..., vT) and (V1, V2, ..., VT), respectively. When calculating the change trend of the foreign exchange transaction volume, the ratio of the real value to the predicted value of the corresponding business day can be calculated, and the ratio corresponding to each business day is used as the change trend of the foreign exchange transaction volume. That is, the change trend of the foreign exchange transaction volume is (V1 / v1, V2 / v2, ..., VT / vT).

The foreign exchange transaction volume in the future time period is predicted according to the change trend, and the foreign exchange transaction volume in the i-th business day can be predicted according to the change trend of multiple business days before the i-th business day. The change trend of multiple business days before 1 business day predicts the foreign exchange transaction volume of the i + 1th business day, and so on. The predicted value of the foreign exchange transaction volume vi (1 ≦ i ≦ T) for each business day can be obtained by performing

steps

104 and 106.

In step 104, the current forecast time node can be set according to actual needs. Generally, the foreign exchange transaction volume before the current forecast time node can be used to predict the foreign exchange transaction volume on the Nth day after the current forecast time node. Assume that the current prediction time node is T, and the foreign exchange transaction volume to be predicted is the foreign exchange transaction volume on the Nth day after the current prediction time node. This prediction method can be called T + N prediction.

Repeat steps 104 and 106 to obtain the T + N forecast result of the current forecast time node T according to the current forecast time node T and the current forecast time node T + N to obtain the current forecast time T + N prediction result at prediction time node T + 1, and so on.

Forex trading volume can be used to generate time series data in chronological order. The foreign exchange transaction volume here can be either the forward transaction volume or the reverse transaction volume. In the consumption scenario, the user's purchase and payment behavior and the user's refund behavior bring about the opposite flow of funds. When the funds are settled, from the perspective of the merchant, the user's payment behavior is a forward transaction, and the user's refund behavior is a Reverse transactions, therefore, the forward transaction volume is the total amount of the user's payment to the merchant, and the reverse transaction volume is the total amount of the user's refund from the merchant.

A business day date attribute and / or a business day promotion attribute of each business day may also be obtained for use in correcting the prediction result of a specific business day. Among them, the business day date attribute is whether the date corresponding to the business day is at the beginning, end of the month, middle of the month, working day, non-working day, or holiday. The business day promotion attribute is whether or not there is a promotion activity and the activity level of the promotion activity on the business day. Among them, the activity level is determined by factors such as the discount strength of the promotion activity, the coverage of merchants, and the expected transaction volume. The time series data of one embodiment is shown in FIG. 2.

In one embodiment, when obtaining the current first prediction result corresponding to each prediction model, the historical foreign exchange transaction volume before the current prediction time node may be input into a pre-trained first prediction model to obtain the first prediction corresponding to the first prediction model. Results; input the historical foreign exchange transaction volume of the N business days closest to the current prediction time node into the pre-trained second prediction model to obtain the first prediction result corresponding to the second prediction model; and the historical synchronization period N current prediction time nodes The historical foreign exchange transaction volume on the business day is input to a pre-trained third prediction model to obtain a first prediction result corresponding to the third prediction model.

The first training model obtains the first prediction result according to the historical full data, the second training model obtains the first prediction result according to the data of the last N business days, and the third training model obtains the prediction result according to the historical synchronization data. The three models obtain forecast results from different perspectives. Finally, the forecast results of each model are combined to obtain the forecast value of foreign exchange transaction volume, which can reduce the forecast bias of each model itself and obtain a more stable forecast result.

Wherein, the first prediction model may be ARIMA (Autoregressive Integrated Moving Model), the second prediction model may be a mean estimation model, and the third prediction model may be a ring-on-trend trend estimation model. . The ARIMA model can sense the periodic trend, growth trend, and seasonal trend of the data. The mean value estimation model is based on the recent data mean, and the forecast value is focused on the imitation of recent stable trends. Trends over the same period in history Predict trends in the current period, with emphasis on imitating historical periods. Of course, the types and number of prediction models used in practical applications are not limited to the above embodiments, and other types of prediction models that can obtain stability trends may be used instead of the above models. An overall schematic diagram of the foreign exchange transaction volume prediction of one embodiment is shown in FIG. 3.

For the ARIMA model, the entire historical data can be directly input to the model to obtain the first prediction result. For the average estimation model, the historical foreign exchange transaction volume of the N business days closest to the current forecast time node can be averaged, and the average value can be used as the first prediction result. For the ring-by-trend trend estimation model, the historical value of the historical period of the current forecast time node (denoted as day1) and the sum of the historical values of day M and day M after day1 can be used. The historical value of the first N days of the current prediction time node is used to divide the historical value of the first N days of day1 to obtain the ring ratio r, and then this ring ratio r is multiplied by the sum of the historical values of M days after day1 to obtain the final prediction. value. The value of M is a positive integer, and optionally, the value of M is 1.

In step 106, the weighted average value of each current first prediction result may be used as the predicted value of the foreign exchange transaction volume on the Nth business day after the current prediction time node. The weight of each first prediction result may be a normalized weight, that is, the sum of the weights is 1.

In one embodiment, the method further comprises: calculating an impact factor of the specific business day according to the predicted value of the foreign exchange transaction volume on the specific business day; and predicting the foreign exchange transaction volume of the specific business day in the future time period according to the impact factor. Value is corrected.

Among them, a specific business day is a business day in which there may be an event that has a large impact on the foreign exchange transaction volume. For example, it can be a business day at the beginning of the month, a business day at the end of the month, or a holiday, or a business day where a big promotion event is held on Double Eleven, Double Twelve, and so on. By correcting the forecast value of the foreign exchange transaction volume on a specific business day, the forecast accuracy can be improved.

Specifically, the ratio of the predicted value of the foreign exchange transaction volume on the specific business day to the actual value of the foreign exchange transaction volume on the specific business day may be used as the influence factor, and the influence factor may be used to determine the specific business in the future time period. Daily foreign exchange transaction volume forecasts are revised.

The method for obtaining the predicted value of the foreign exchange transaction volume on a specific business day may be similar to step 104. That is, the foreign exchange transaction volume of a specific business day is input into the multiple prediction models, and second prediction results corresponding to each prediction model are obtained; and the predicted value of the foreign exchange transaction volume of the specific business day is calculated according to each second prediction result. Optionally, the multiple prediction models used here may be the same as the prediction models used in step 104, and the prediction mode of each model may also be the same as the prediction mode of each prediction model in step 104.

Further, after obtaining the second prediction result corresponding to each prediction model, the weighted average value of each second prediction result may be used as the foreign exchange transaction volume prediction value on the specific business day.

In the embodiment of the present specification, for a time series without more information, when only the date attributes of a specific business day are known, the N-day data prediction of T + N in a specific business day is divided into two parts, one part is the stable trend of the sequence. One part is due to the influencing factors and trends brought by a specific business day. Finally, the two parts of the trend are merged to obtain the forecast value of the foreign exchange transaction volume in the future time period, which improves the accuracy of the forecast results.

The various technical features in the above embodiments can be arbitrarily combined, as long as there is no conflict or contradiction in the combination of features, but it is limited in space and has not been described one by one, so the various technical features in the above embodiments are arbitrarily combined. It also belongs to the scope disclosed in this specification.

As shown in FIG. 4, an embodiment of the present specification further provides a device for predicting foreign exchange transaction volume. The device may include:

The prediction module 402 is configured to calculate a change trend of the foreign exchange transaction volume according to the predicted value of the foreign exchange transaction volume of each business day and the real value of the foreign exchange transaction volume corresponding to the business day, and to predict the foreign exchange transaction volume in the future time period according to the change trend. Perform forecasting, where the forecast value of foreign exchange transaction volume for each business day is obtained by executing the functions of the following modules:

The input module 404 is configured to input the historical foreign exchange transaction volume before the current prediction time node into a plurality of pre-trained prediction models to obtain the current first prediction result corresponding to each prediction model;

The calculation module 406 is configured to calculate the predicted value of the foreign exchange transaction amount on the Nth business day after the current prediction time node according to each current first prediction result; N is a preset positive integer.

For details of the implementation process of the functions and functions of each module in the above device, refer to the implementation process of corresponding steps in the foregoing method, and details are not described herein again.

As for the device embodiment, since it basically corresponds to the method embodiment, the relevant part may refer to the description of the method embodiment. The device embodiments described above are only schematic, and the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, which may be located in One place, or can be distributed to multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution in this specification. Those of ordinary skill in the art can understand and implement without creative efforts.

The embodiments of the apparatus of the present specification can be applied to a computer device, such as a server or a terminal device. The device embodiments may be implemented by software, or by hardware or a combination of software and hardware. Taking software implementation as an example, as a device in a logical sense, it is formed by reading the corresponding computer program instructions in the non-volatile memory into the memory and running the processor through the file processing processor. In terms of hardware, as shown in FIG. 5, it is a hardware structure diagram of the computer equipment where the device of this specification is located, except for the processor 502, the memory 504, the network interface 506, and the non-volatile memory 508 shown in FIG. 5. In addition, the server or electronic device where the device is located in the embodiment may generally include other hardware according to the actual function of the computer device, and details are not described herein again.

Correspondingly, the embodiment of the present specification also provides a computer storage medium. The storage medium stores a program, and when the program is executed by a processor, the method in any of the foregoing embodiments is implemented.

Accordingly, an embodiment of the present specification further provides a computer device including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, any one of the foregoing embodiments is implemented. Method.

This application may take the form of a computer program product implemented on one or more storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing program code. Computer-usable storage media includes permanent and non-permanent, removable and non-removable media, and information can be stored by any method or technology. Information may be computer-readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to: phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, read-only disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, Magnetic tape cartridges, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media may be used to store information that can be accessed by computing devices.

Those skilled in the art will readily think of other embodiments of the present disclosure after considering the specification and practicing the specification disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include the common general knowledge or conventional technical means in the technical field not disclosed by this disclosure. . It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

It should be understood that the present disclosure is not limited to the precise structure that has been described above and illustrated in the drawings, and various modifications and changes can be made without departing from the scope thereof. The scope of the disclosure is limited only by the following claims.

The above are merely preferred embodiments of the present disclosure, and are not intended to limit the present disclosure. Any modification, equivalent replacement, or improvement made within the spirit and principles of the present disclosure shall be included in the present disclosure. Within the scope of protection.

Claims

A method for predicting foreign exchange transaction volume, the method includes:

Calculate the change trend of foreign exchange transaction volume based on the predicted value of foreign exchange transaction volume for each business day and the real value of foreign exchange transaction volume for the corresponding business day, and predict the foreign exchange transaction volume in the future time period based on the change trend, where each The forecast value of foreign exchange transaction volume on the business day is obtained according to the following methods:

Input the historical foreign exchange transaction volume before the current prediction time node into multiple pre-trained prediction models to obtain the current first prediction result corresponding to each prediction model;

Calculate the predicted value of foreign exchange transaction volume on the Nth business day after the current prediction time node according to each current first prediction result; N is a preset positive integer.
The method according to claim 1, wherein the step of calculating the predicted value of the foreign exchange transaction volume on the Nth business day after the current prediction time node according to each current first prediction result comprises:

The weighted average value of each current first prediction result is used as the predicted value of the foreign exchange transaction volume on the Nth business day after the current prediction time node.
The method according to claim 1, further comprising:

Calculating the impact factor of the specific business day according to the forecast value of the foreign exchange transaction volume on the specific business day;

Correct the predicted value of the foreign exchange transaction volume for a specific business day in the future time period according to the impact factor.
The method according to claim 3, the step of calculating the impact factor of the specific business day based on the predicted value of foreign exchange transaction volume on the specific business day comprises:

The ratio of the predicted value of the foreign exchange transaction volume on the specific business day to the true value of the foreign exchange transaction volume on the specific business day is used as the influence factor.
The method according to claim 3, further comprising:

Input the foreign exchange transaction volume of a specific business day into the plurality of prediction models, and obtain second prediction results corresponding to each prediction model;

Calculate the predicted foreign exchange transaction volume for the specific business day according to each second prediction result.
The method according to claim 5, wherein the step of calculating a forecast value of the foreign exchange transaction volume on the specific business day according to each second prediction result comprises:

The weighted average of each second prediction result is used as the predicted value of the foreign exchange transaction volume on the specific business day.
The method according to any one of claims 1 to 6, respectively inputting a historical foreign exchange transaction volume before a current prediction time node into a plurality of pre-trained prediction models, and obtaining the current first prediction result corresponding to each prediction model includes:

Input the historical foreign exchange transaction volume before the current prediction time node into a pre-trained first prediction model to obtain a first prediction result corresponding to the first prediction model;

Input the historical foreign exchange transaction volume of the N business days closest to the current prediction time node into a pre-trained second prediction model to obtain a first prediction result corresponding to the second prediction model; and

The historical foreign exchange transaction volume of the N historical business days during the current forecast time node is input into a pre-trained third prediction model to obtain a first prediction result corresponding to the third prediction model.
The method according to claim 7, wherein the first prediction model is an ARIMA model, the second prediction model is a mean estimation model, and the third prediction model is a chain-wise trend estimation model.
A foreign exchange transaction volume prediction device, the device includes:

The forecasting module is configured to calculate a change trend of the foreign exchange transaction volume according to the predicted value of the foreign exchange transaction volume of each business day and the real value of the foreign exchange transaction volume of the corresponding business day, and perform a foreign exchange transaction volume in the future time period according to the change trend Forecast, in which the forecast value of foreign exchange transaction volume for each business day is obtained by executing the functions of the following modules:

An input module for inputting historical foreign exchange transaction volumes before the current prediction time node into a plurality of pre-trained prediction models to obtain the current first prediction result corresponding to each prediction model;

A calculation module is configured to calculate a predicted value of a foreign exchange transaction amount on an N-th business day after the current prediction time node according to each current first prediction result; N is a preset positive integer.
A computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 8.
A computer device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the method according to any one of claims 1 to 8 is implemented.