Training method, prediction method, device, computer system and storage medium

Abstract

The present disclosure provides a method for training a prediction model, which can be used in the financial field, the artificial intelligence field or the computer technology field. The training method of the prediction model includes: obtaining training samples, the training samples include training data and training labels corresponding to the training data, the training data includes data related to the handling of business, and the training labels include the historical business processing volume of the target business handled within the target time period; obtaining multiple initial models, wherein the network structures of the multiple initial models are different; using the training samples to train the multiple initial models to obtain multiple prediction models, each prediction model corresponds to an initial model in the multiple initial models; and determining the target prediction model from the multiple prediction models. The present disclosure also provides a business processing volume prediction method, a prediction model training device, a business processing volume prediction device, a computer system, a readable storage medium and a computer program product.

Description

Technical Field

[0001] The present disclosure relates to the fields of finance, artificial intelligence or computer technology, and more specifically, to a training method, a prediction method, a device, a computer system, a storage medium and a computer program product. Background Art

[0002] Business outlets are the main places where users conduct business. For example, bank outlets handle various financial services such as deposits and withdrawals, loans, or purchase of financial products for users. They can also be business outlets in other industries, such as business outlets in hospitals and airports, which can handle consulting services, payment services, etc. for users.

[0003] In the process of realizing the concept of the present disclosure, the inventors discovered that there are at least the following problems in the related art: the business volume undertaken by the business outlets cannot be reasonably matched with the actual business volume handled. Summary of the invention

[0004] In view of this, the present disclosure provides a prediction model training method, a business processing volume prediction method, an apparatus, a computer system, a readable storage medium, and a computer program product.

[0005] One aspect of the present disclosure provides a method for training a prediction model, comprising:

[0006] Acquire training samples, wherein the training samples include training data and training labels corresponding to the training data, wherein the training data includes data related to handling business, and the training labels include historical business processing volume of handling target business within a target time period;

[0007] Acquire multiple initial models, wherein each of the multiple initial models has a different network structure;

[0008] Using the training samples to train the multiple initial models to obtain multiple prediction models, wherein each prediction model corresponds to an initial model among the multiple initial models; and

[0009] A target prediction model is determined from the multiple prediction models mentioned above.

[0010] According to an embodiment of the present disclosure, the training data includes at least one of the following:

[0011] Weather data, date data, time data, geographic location data, business type data.

[0012] According to an embodiment of the present disclosure, the initial model includes at least one of the following:

[0013] Autoregressive moving average model, autoregressive difference moving average model, autoregressive model, moving average model, long short-term memory network model.

[0014] According to an embodiment of the present disclosure, determining the target prediction model from the above-mentioned multiple prediction models includes:

[0015] Acquire a test sample, wherein the test sample includes test data and a test tag corresponding to the test data, wherein the test data includes data related to handling of a business, and the test tag includes a historical business processing volume of handling a target business within a target time period;

[0016] Inputting the test data into each of the plurality of prediction models to obtain the business processing volume test results of each of the plurality of prediction models; and

[0017] Based on the business processing volume test results of each of the above-mentioned multiple prediction models and the above-mentioned test labels, the above-mentioned target prediction model is determined.

[0018] Another aspect of the present disclosure provides a method for predicting a service processing volume, comprising:

[0019] Obtaining characteristic data related to handling the business to be predicted within the time period to be predicted; and

[0020] Input the above characteristic data into the target prediction model to obtain the business processing volume prediction result of the above business to be predicted within the above time period to be predicted;

[0021] The target prediction model is trained using the prediction model training method.

[0022] According to an embodiment of the present disclosure, the above-mentioned target prediction model includes multiple;

[0023] The above-mentioned characteristic data is input into the target prediction model to obtain the business processing volume prediction result of the above-mentioned business to be predicted in the above-mentioned prediction time period, including:

[0024] Inputting the characteristic data into a plurality of target prediction models to obtain the business processing volume prediction results of the plurality of target prediction models for the business to be predicted within the time period to be predicted;

[0025] Based on the business processing volume forecast results of the above-mentioned businesses to be predicted within the above-mentioned time period to be predicted of each of the above-mentioned multiple target forecasting models, the target processing volume forecast results are determined.

[0026] According to an embodiment of the present disclosure, the above-mentioned service processing volume prediction method further includes:

[0027] Obtaining the actual results of the business processing volume of the above-mentioned predicted business within the predicted time period; and

[0028] Based on the business processing volume prediction results of the above-mentioned businesses to be predicted within the above-mentioned time period and the actual results of the above-mentioned business processing volume of each of the above-mentioned multiple target prediction models, a final prediction model is determined from the above-mentioned multiple target prediction models, wherein the above-mentioned final prediction model includes one.

[0029] Another aspect of the present disclosure provides a training device for a prediction model, comprising:

[0030] A first acquisition module is used to acquire training samples, wherein the training samples include training data and training labels corresponding to the training data, wherein the training data includes data related to handling business, and the training labels include historical business processing volume of handling target business within a target time period;

[0031] A second acquisition module is used to acquire multiple initial models, wherein the network structures of the multiple initial models are different;

[0032] a training module, configured to train the plurality of initial models using the training samples to obtain a plurality of prediction models, wherein each of the prediction models corresponds to an initial model among the plurality of initial models; and

[0033] The first determination module is used to determine a target prediction model from the above-mentioned multiple prediction models.

[0034] Another aspect of the present disclosure provides a service processing volume prediction device, comprising:

[0035] A third acquisition module is used to acquire characteristic data related to the business to be predicted within the time period to be predicted; and

[0036] The second determination module is used to input the above-mentioned characteristic data into the target prediction model to obtain the business processing volume prediction result of the above-mentioned business to be predicted within the above-mentioned prediction time period;

[0037] The target prediction model is trained using the prediction model training method.

[0038] Another aspect of the present disclosure provides a computer system, comprising:

[0039] one or more processors;

[0040] a memory for storing one or more programs,

[0041] Among them, when the above-mentioned one or more programs are executed by the above-mentioned one or more processors, the above-mentioned one or more processors implement the above-mentioned prediction model training method, or implement the above-mentioned business processing volume prediction method.

[0042] Another aspect of the present disclosure provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enables the processor to implement the above-mentioned prediction model training method, or implement the above-mentioned business processing volume prediction method.

[0043] Another aspect of the present disclosure provides a computer program product, which includes computer executable instructions. When the instructions are executed, they are used to implement the training method of the prediction model or to implement the business processing volume prediction method.

[0044] According to the embodiments of the present disclosure, because a training sample is obtained, wherein the training sample includes training data and training labels corresponding to the training data, wherein the training data includes data related to handling business, and the training labels include the historical business processing volume of handling the target business within the target time period; multiple initial models are obtained, wherein the network structures of the multiple initial models are different; multiple initial models are trained using the training samples to obtain multiple prediction models, wherein each prediction model corresponds to one of the multiple initial models; the technical means for determining the target prediction model from multiple prediction models can be used to train multiple different initial models using the training samples, and the multiple prediction models obtained can be used to determine the target prediction model based on the multiple prediction models. In the case of uncertainty of the appropriate model architecture, the applicability of the target prediction model and the accuracy of the prediction are improved by trying and considering multiple initial models. Then, the target prediction model can be used to accurately predict the business processing volume. Therefore, the technical problem that the business volume of offline business conditions cannot be analyzed in the related art is at least partially overcome. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The above and other objects, features and advantages of the present disclosure will become more apparent through the following description of the embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0046] Figure 1 The application scenario diagram schematically shows a prediction model training method, a business processing volume prediction method, a prediction model training device, and a business processing volume prediction device according to an embodiment of the present disclosure;

[0047] Figure 2 A flowchart of a method for training a prediction model according to an embodiment of the present disclosure is schematically shown;

[0048] Figure 3A flowchart of determining a target prediction model from multiple prediction models according to an embodiment of the present disclosure is schematically shown;

[0049] Figure 4 A flowchart of a method for predicting business processing volume according to an embodiment of the present disclosure is schematically shown;

[0050] Figure 5 The application flow chart of the business processing volume prediction method according to the embodiment of the present disclosure is schematically shown;

[0051] Figure 6 The structure block diagram of the training device of the prediction model according to the embodiment of the present disclosure is schematically shown;

[0052] Figure 7 The structure block diagram of the service processing volume prediction device according to the embodiment of the present disclosure is schematically shown;

[0053] Figure 8 The structural block diagram of the prediction model training method and the business processing volume prediction method according to the embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0054] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0055] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0056] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0057] In the case of using expressions such as "at least one of A, B, and C, etc.", it should generally be interpreted in accordance with the meaning of the expression generally understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.). In the case of using expressions such as "at least one of A, B, or C, etc.", it should generally be interpreted in accordance with the meaning of the expression generally understood by those skilled in the art (for example, "a system having at least one of A, B, or C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0058] In the related art, a business outlet is the main place where users conduct business. For example, a bank business outlet handles various financial services such as deposit and withdrawal, loan or purchase of financial products for users. It can also be a business outlet in other industries, such as a business outlet in a hospital or airport, which can handle consulting services, payment services, etc. for users.

[0059] In the process of realizing the concept disclosed in the present invention, the inventors found that there are at least the following problems in the related technology: the user flow information of the current business outlets is mainly analyzed through dimensions such as the business processing volume in the system and the number of users taking numbers, and the user flow information, such as the number of users handling business, is mainly analyzed through data retained by the online system, and there is no effective statistical and management method for offline business conditions. Therefore, it is impossible to analyze the offline business conditions and business volume, such as the number of users who have not taken numbers, the number of business consultations, the number of product recommendations, the number of users at different windows, etc., and it is impossible to predict a surge or decrease in user flow in future time periods.

[0060] In order to at least partially solve the technical problems existing in the related technologies, the present disclosure provides a prediction model training method and a business processing volume prediction method, which can be applied to the financial field, the artificial intelligence technology field or the computer technology field. For example, the prediction model training method and the business processing volume prediction method are applied in a bank branch to predict the business processing volume of the bank branch, or the business processing volume of multiple bank branches can be predicted. It can also be used in any field other than the financial field, the artificial intelligence technology field or the computer technology field. The method, device, computer system, readable storage medium and computer program product provided by the present disclosure do not limit the application field.

[0061] In the technical solution disclosed in the present invention, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations, necessary confidentiality measures are taken, and do not violate public order and good morals.

[0062] The embodiment of the present disclosure provides a prediction model training method, comprising:

[0063] Obtain training samples, wherein the training samples include training data and training labels corresponding to the training data, wherein the training data includes data related to handling business, and the training labels include the historical business processing volume of handling target business within the target time period; obtain multiple initial models, wherein the network structures of the multiple initial models are different; use the training samples to train the multiple initial models to obtain multiple prediction models, wherein each prediction model corresponds to one of the multiple initial models; and determine the target prediction model from the multiple prediction models. Determine the target prediction model from the multiple prediction models. The technical means for determining the target prediction model from the multiple prediction models can use the training samples to train multiple different initial models, and obtain multiple prediction models, and determine the target prediction model based on the multiple prediction models. In the case of uncertainty of the suitable model architecture, by trying and considering multiple initial models, the applicability of the target prediction model and the accuracy of the prediction are improved. Then, the target prediction model can be used to accurately predict the business processing volume. Therefore, at least part of the technical problem that the offline business situation and business volume cannot be analyzed in the related art is overcome.

[0064] The embodiment of the present disclosure also provides a method for predicting service processing volume, including:

[0065] Acquire characteristic data related to the handling of the business to be predicted within the time period to be predicted; and input the characteristic data into the target prediction model to obtain the business processing volume prediction result of the business to be predicted within the time period to be predicted; wherein the target prediction model is trained using the training method of the above prediction model.

[0066] Therefore, the target prediction model is trained by the above-mentioned prediction model training method, and the final prediction model is determined based on the comparison between the business processing volume prediction results and the actual results. Compared with using only one prediction model to predict the business processing volume, the accuracy of the prediction results can be further improved.

[0067] Figure 1 The following schematically shows the application scenario diagram of the prediction model training method, the business processing volume prediction method, the prediction model training device, and the business processing volume prediction device according to the embodiment of the present disclosure. It should be noted that: Figure 1 What is shown is merely an example of a system architecture to which the embodiments of the present disclosure can be applied, in order to help those skilled in the art understand the technical content of the present disclosure, but it does not mean that the embodiments of the present disclosure cannot be used in other devices, systems, environments or scenarios.

[0068] like Figure 1As shown, the system architecture 100 according to this embodiment may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired and / or wireless communication links, etc.

[0069] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, 103, such as video data processing applications, web browser applications, search applications, instant messaging tools, email clients and / or social platform software, etc. (only as examples).

[0070] The terminal devices 101 , 102 , and 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, and desktop computers.

[0071] The server 105 may be a server that provides various services, such as a background management server (only as an example) that provides support for websites browsed by users using the terminal devices 101, 102, and 103. The background management server may analyze and process the received user requests and other data related to the handling of services, and feed back the processing results (such as web pages, information, or data obtained or generated according to the user requests) to the terminal device.

[0072] It should be noted that the prediction model training method and the business processing volume prediction method provided in the embodiment of the present disclosure can generally be executed by the server 105. Accordingly, the business processing volume prediction device provided in the embodiment of the present disclosure can generally be set in the server 105. The prediction model training method and the business processing volume prediction method provided in the embodiment of the present disclosure can also be executed by a server or server cluster that is different from the server 105 and can communicate with the terminal devices 101, 102, 103 and / or the server 105. Accordingly, the business processing volume prediction device provided in the embodiment of the present disclosure can also be set in a server or server cluster that is different from the server 105 and can communicate with the terminal devices 101, 102, 103 and / or the server 105.

[0073] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0074] Figure 2 A flowchart of a method for training a prediction model according to an embodiment of the present disclosure is schematically shown.

[0075] like Figure 2 As shown, the prediction model training method of this embodiment includes operations S201 to S204.

[0076] In operation S201, a training sample is obtained, wherein the training sample includes training data and training labels corresponding to the training data, wherein the training data includes data related to handling business, and the training labels include historical business processing volume of handling target business within a target time period.

[0077] According to an embodiment of the present disclosure, the training sample may be a first sample set selected from a historical sample set. The historical samples in the historical sample set may be data related to business handling and actual business processing volume in a historical period. It is worth noting that the data related to business handling and actual business processing volume in the historical samples may be pre-processed, such as filtering noise data, etc., in order to improve data accuracy.

[0078] According to an embodiment of the present disclosure, historical samples may be obtained by querying business processing information in the system within a historical period.

[0079] According to an embodiment of the present disclosure, the training data may be, for example, data related to business handling in historical time. The training label may be the actual business processing volume in the historical time corresponding to the training data.

[0080] According to the embodiments of the present disclosure, the historical time may be a time period before the current time based on the current time. Those skilled in the art may design the historical time according to actual needs.

[0081] According to an embodiment of the present disclosure, the target time period may be one or more time periods in historical time, for example, it may be 9:00 to 11:00 on X / X / 2019, or it may be 9:00 to 11:00 on X / X / 2019 and 9:00 to 11:00 on X / Y / 2019. Those skilled in the art may set the target time period according to actual needs.

[0082] According to the embodiments of the present disclosure, the target business may be, for example, a deposit and withdrawal business, a loan business, or a payment business, but is not limited thereto and may also be other businesses.

[0083] According to the embodiments of the present disclosure, the historical business processing volume can be the number of businesses that process the target business in a certain business outlet, or the number of businesses that process the target business in multiple business outlets. The number of users who process the business can be obtained based on the number of businesses that process the target business.

[0084] It should be noted that the benchmark method for historical business volume is not limited to the method disclosed above in the present invention, and can also be set by technical personnel based on technical requirements.

[0085] In operation S202, a plurality of initial models are obtained, wherein the network structures of the plurality of initial models are different.

[0086] According to an embodiment of the present disclosure, the initial model may be constructed based on time series analysis, but is not limited thereto, and may also be constructed based on other machine learning models.

[0087] In operation S203, a plurality of initial models are trained using the training samples to obtain a plurality of prediction models, wherein each prediction model corresponds to an initial model among the plurality of initial models.

[0088] According to an embodiment of the present disclosure, training the initial model can be achieved by adjusting the parameters in the initial model. For example, when the initial model is a long short-term memory network model (LSTM model) constructed based on a machine learning model, the training result can be obtained based on the training data and the initial model, and the training result can be compared with the training label to obtain the comparison result, and the parameters of the initial model can be adjusted based on the comparison result. Repeat the above process, and when the comparison result is less than or equal to the preset training threshold, stop the training and obtain the prediction model.

[0089] But it is not limited to this. When the initial model is an autoregressive moving average model (ARMA model) constructed based on time series analysis, the initial model can be trained by testing the residuals. For example, when the model residuals obtained by testing the residuals are white noise with a normal distribution, a prediction model is obtained.

[0090] In operation S204, a target prediction model is determined from among the plurality of prediction models.

[0091] According to an embodiment of the present disclosure, a target prediction model can be determined based on the test results of each prediction model in a plurality of prediction models. For example, a second sample set can be selected from a historical sample set, and the historical samples in the second sample set can be used as test samples, and the test results of each prediction model can be obtained based on the test samples and the plurality of prediction models, and then the test results of each prediction model can be compared, and the target prediction model can be determined based on the comparison results.

[0092] According to the embodiments of the present disclosure, the technical means for determining the target prediction model from multiple prediction models can use training samples to train multiple different initial models, and obtain multiple prediction models, and determine the target prediction model based on the multiple prediction models. In the case of uncertainty about the appropriate model architecture, the applicability of the target prediction model and the accuracy of the prediction are improved by trying and considering multiple initial models. The target prediction model can then be used to accurately predict the business processing volume. Therefore, the technical problem in the related art that the offline business situation and business volume cannot be analyzed is at least partially overcome.

[0093] According to an embodiment of the present disclosure, the training data may include at least one of weather data, date data, time data, geographic location data, or business category data.

[0094] According to an embodiment of the present disclosure, weather data may include weather conditions, such as sunny, cloudy, hail, and rainstorm, but is not limited thereto, and may also include other weather data such as temperature, air pressure, humidity, and wind level.

[0095] According to an embodiment of the present disclosure, date data may include, for example, a specific date in the Gregorian calendar, the lunar calendar, or other calendars, and may also include attribute information of the date data. The attribute information may include, for example, holiday attribute information such as weekends and Labor Day holidays, and may also include financial attribute information such as pension issuance, bond issuance, and commemorative coin issuance, but is not limited thereto, and may also include other attribute information of the date data.

[0096] According to an embodiment of the present disclosure, time data may include one or more time intervals within a business day, such as 9:00 to 11:00, 11:00 to 13:00, but is not limited to this, and may also be one or more time points. Those skilled in the art can design the time data to be obtained according to actual needs.

[0097] According to an embodiment of the present disclosure, the geographic location data may include a geographic location, such as geographic coordinates, and may also include a preset geographic area containing the geographic coordinates and attribute information of the preset geographic area. The attribute information of the preset geographic area may include the population of the preset geographic area, the per capita annual income of the preset geographic area, and the number of people of a preset age in the preset geographic area. For example, considering that there are a large number of elderly users who go to business outlets to handle business, the number of people of a preset age may be set to: the number of people aged 65 to 70. In addition, the attribute information of the preset geographic area may include the number of users entering the business outlets, such as the number of users who take numbers at self-service terminal devices at bank business outlets.

[0098] According to an embodiment of the present disclosure, business type data may include the business type of the business handled, for example, it may include deposit and withdrawal business, financial management business, insurance business, precious metal business, transfer business, foreign exchange business, and account opening business, but is not limited to this, and may also be other business types, such as payment business and consulting business.

[0099] It should be understood that the handled business may have one business class or multiple business classes, and those skilled in the art may design the business type data to be obtained according to actual needs.

[0100] According to an embodiment of the present disclosure, at least one of weather data, date data, time data, geographic location data or business type data is used as training data. Those skilled in the art may consider the impact of the above multiple factors as training data on the predicted business processing volume to improve the accuracy of the predicted business processing volume.

[0101] According to an embodiment of the present disclosure, the initial model may include: at least one of an autoregressive moving average model (ARMA model), an autoregressive difference moving average model (ARIMA model), an autoregressive model (AR model), a moving average model (MA model), and a long short-term memory network model (LSTM model).

[0102] Figure 3 The flowchart of determining a target prediction model from multiple prediction models according to an embodiment of the present disclosure is schematically shown.

[0103] like Figure 3 As shown, determining a target prediction model from multiple prediction models in this embodiment may include operations S301 to S303.

[0104] In operation S301, a test sample is obtained, wherein the test sample includes test data and a test tag corresponding to the test data, wherein the test data includes data related to handling business, and the test tag includes a historical business processing volume of handling a target business within a target time period.

[0105] According to an embodiment of the present disclosure, the test sample may be a second sample set selected from a historical sample set, and the historical samples in the second sample set are used as the test sample.

[0106] According to an embodiment of the present disclosure, the test data may be, for example, data related to business handling in a historical period, and the test tag may be the actual business processing volume in the historical period corresponding to the test data.

[0107] According to an embodiment of the present disclosure, the test data may include at least one of weather data, date data, time data, geographic location data, or business category data.

[0108] In operation S302, test data is input into each prediction model of the plurality of prediction models to obtain business processing volume test results of each of the plurality of prediction models.

[0109] According to an embodiment of the present disclosure, the business processing volume test result is a prediction result of each prediction model for the business processing volume of the target business within the target time period.

[0110] In operation S303, a target prediction model is determined based on the business processing volume test results and test labels of each of the multiple prediction models.

[0111] According to an embodiment of the present disclosure, for example, the business processing volume test results of multiple prediction models can be compared with the test labels to determine whether they are less than or equal to a preset comparison threshold. If the judgment result is yes, the prediction model corresponding to the business processing volume test result is determined as the target prediction model.

[0112] According to an embodiment of the present disclosure, by comparing the business processing volume test results of each of a plurality of prediction models with the test labels and determining the target prediction model based on the comparison results, at least the prediction models whose business processing volume test results are significantly different from the actual business processing volume results can be screened out, that is, the prediction models with poor prediction effects can be screened out, thereby enabling technical personnel in the field to use the target prediction model to predict business volume to improve the accuracy of the prediction.

[0113] Figure 4 The flowchart of the business processing volume prediction method according to the embodiment of the present disclosure is schematically shown.

[0114] like Figure 4 As shown, the service processing volume prediction method of this embodiment includes operations S401-S402.

[0115] In operation S401, characteristic data related to handling of the to-be-forecasted business within the to-be-forecasted time period is obtained.

[0116] According to an embodiment of the present disclosure, feature data may be data having the same attributes as training data and test data, for example, may include at least one of weather data, date data, time data, geographic location data or business category data.

[0117] According to the embodiments of the present disclosure, the time period to be predicted may be a time period after the current time based on the current time. It may also refer to a time period after the current date based on the current date; it may also refer to a time period after the current year based on the current year. It should be noted that the reference method for the time period to be predicted is not limited to the method disclosed above in the present invention, and can also be set by the technicians based on technical needs.

[0118] In operation S402, the characteristic data is input into the target prediction model to obtain a prediction result of the business processing volume of the predicted business within the predicted time period; wherein the target prediction model is trained using the training method of the above prediction model.

[0119] According to an embodiment of the present disclosure, the target prediction model may be one or more.

[0120] According to the embodiments of the present disclosure, the target prediction model is a prediction model that screens out prediction models with poor prediction effects. Since there can be multiple target prediction models, technicians in this field can select one or more target prediction models for predicting the business processing volume of the business to be predicted within the prediction time period based on the prediction results of each prediction model in the multiple target prediction models, the model operation speed and other factors.

[0121] According to the embodiments of the present disclosure, those skilled in the art can divide a plurality of time periods to be predicted, for example, they can easily divide them into time period A to be predicted, time period B to be predicted, and time period C to be predicted in chronological order, and use different target prediction models to predict the business processing volume in each time period to be predicted, so that those skilled in the art can select a suitable target prediction model to predict the business processing volume based on the attribute information of the feature data in the time period to be predicted.

[0122] According to an embodiment of the present disclosure, the target prediction model may include multiple ones.

[0123] Inputting the characteristic data into the target prediction model, the business processing volume prediction results of the predicted business in the predicted time period may include:

[0124] The characteristic data is input into multiple target prediction models to obtain the business processing volume prediction results of the multiple target prediction models for the businesses to be predicted within the respective prediction time periods; based on the business processing volume prediction results of the multiple target prediction models for the businesses to be predicted within the respective prediction time periods, the target processing volume prediction results are determined.

[0125] According to an embodiment of the present disclosure, the service processing volume prediction method of this embodiment may further include:

[0126] Obtain the actual results of the business processing volume of the business to be predicted within the time period to be predicted; based on the business processing volume prediction results and actual results of the business processing volume of the business to be predicted within the time period to be predicted of multiple target prediction models, determine the final prediction model from multiple target prediction models, wherein the final prediction model includes one.

[0127] According to an embodiment of the present disclosure, the business processing volume prediction results of multiple target prediction models for the businesses to be predicted within the prediction time period can be compared with the actual business processing volume results to determine whether they are less than or equal to a preset determination threshold. If the judgment result is yes, the target prediction model corresponding to the business processing volume prediction result will be determined as the final prediction model.

[0128] It should be understood that the final prediction model can be determined as one by setting a preset determination threshold.

[0129] According to an embodiment of the present disclosure, the preset determination threshold may be, for example, the difference between the predicted result of the business processing volume and the actual result of the business processing volume, but is not limited thereto. Those skilled in the art may design the preset determination threshold according to actual needs.

[0130] According to other embodiments of the present disclosure, the business processing volume prediction results can also be compared with the actual business processing volume results to obtain the prediction accuracy of each target prediction model, and then the target prediction models are sorted according to the accuracy, and the target prediction model with the highest accuracy is determined as the final prediction model.

[0131] According to an embodiment of the present disclosure, multiple business processing volume prediction results are generated based on multiple target prediction models, and the closest business processing volume prediction result is determined based on the multiple business processing volume prediction results and the actual business processing volume results, and the target prediction model corresponding to the closest business processing volume prediction result is determined as the final prediction model. Compared with the relevant existing technology that always uses only one prediction model to predict the business processing volume, technical personnel in this field can select the final prediction model suitable for predicting the business processing volume according to the actual needs in the application scenario, thereby further improving the accuracy of the prediction results.

[0132] Figure 5 The application flow chart of the business processing volume prediction method according to an embodiment of the present disclosure is schematically shown.

[0133] According to the embodiments of the present disclosure, Figure 5 As shown, the business processing volume prediction method in this embodiment can be implemented through the following process.

[0134] Acquire feature data 501, input feature data 501 into target prediction model A, target prediction model B, and target prediction model C, and obtain prediction results A, B, and C, respectively. Compare prediction results A, B, and C with actual results of business processing volume, respectively, to determine whether they are less than or equal to a preset determination threshold.

[0135] For example, if prediction result B is less than or equal to a preset determination threshold, and prediction results A and C are greater than the preset determination threshold, target prediction model B may be determined as the final prediction model 502. Target prediction model A and target prediction model C are eliminated.

[0136] Figure 6 The structural block diagram of a training device for a prediction model according to an embodiment of the present disclosure is schematically shown.

[0137] like Figure 6 As shown, the prediction model training device 600 of this embodiment includes a first acquisition module 601 , a second acquisition module 602 , a training module 603 and a first determination module 604 .

[0138] The first acquisition module 601 is used to acquire training samples, wherein the training samples include training data and training labels corresponding to the training data, wherein the training data includes data related to handling business, and the training labels include historical business processing volume of handling target business within a target time period.

[0139] The second acquisition module 602 is used to acquire multiple initial models, wherein the network structures of the multiple initial models are different.

[0140] The training module 603 is used to train multiple initial models using training samples to obtain multiple prediction models, wherein each prediction model corresponds to an initial model among the multiple initial models.

[0141] The first determination module 604 is used to determine a target prediction model from multiple prediction models.

[0142] According to an embodiment of the present disclosure, the training data includes at least one of weather data, date data, time data, geographic location data, and business category data.

[0143] According to an embodiment of the present disclosure, the initial model includes at least one of an autoregressive moving average model, an autoregressive difference moving average model, an autoregressive model, a moving average model, and a long short-term memory network model.

[0144] According to an embodiment of the present disclosure, the first determination module 604 includes: a first acquisition unit, a first prediction unit and a first determination unit.

[0145] The first acquisition unit is used to acquire a test sample, wherein the test sample includes test data and a test tag corresponding to the test data, wherein the test data includes data related to handling business, and the test tag includes a historical business processing volume of handling a target business within a target time period.

[0146] The first prediction unit is used to input test data into each prediction model of the multiple prediction models to obtain the business processing volume test results of each of the multiple prediction models.

[0147] The first determination unit is used to determine a target prediction model based on the business processing volume test results and test labels of each of the multiple prediction models.

[0148] According to an embodiment of the present disclosure, any multiple modules of the first acquisition module 601, the second acquisition module 602, the training module 603 and the first determination module 604 can be combined into one module for implementation, or any one of the modules can be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the first acquisition module 601, the second acquisition module 602, the training module 603 and the first determination module 604 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation methods of software, hardware and firmware or in an appropriate combination of any of them. Alternatively, at least one of the first acquisition module 601, the second acquisition module 602, the training module 603 and the first determination module 604 may be at least partially implemented as a computer program module, and when the computer program module is executed, the corresponding function may be executed.

[0149] Figure 7 The structural block diagram of the business processing volume prediction device according to an embodiment of the present disclosure is schematically shown.

[0150] like Figure 7 As shown, the service processing volume prediction device 700 of this embodiment includes: a third acquisition module 701 and a second determination module 702

[0151] The third acquisition module 701 is used to acquire characteristic data related to the business to be predicted within the time period to be predicted; and

[0152] The second determination module 702 is used to input the characteristic data into the target prediction model to obtain the prediction result of the business processing volume of the predicted business in the predicted time period. The target prediction model is trained using the training method of the prediction model in the above embodiment.

[0153] According to an embodiment of the present disclosure, the target prediction model may include multiple ones. The second determination module 702 includes: a second prediction unit and a second determination unit.

[0154] The second prediction unit is used to input the characteristic data into multiple target prediction models to obtain the business processing volume prediction results of the multiple target prediction models for the businesses to be predicted within the respective prediction time periods.

[0155] The second determination unit is used to determine the target processing volume prediction result based on the business processing volume prediction results of the to-be-predicted business within the to-be-predicted time period of each of the multiple target prediction models.

[0156] According to an embodiment of the present disclosure, the service processing volume prediction device 700 of this embodiment further includes:

[0157] The fourth acquisition module is used to obtain the actual result of the business processing volume of the predicted business within the predicted time period.

[0158] The third determination module is used to determine a final prediction model from multiple target prediction models based on the business processing volume prediction results and actual business processing volume results of the businesses to be predicted within the respective prediction time periods of the multiple target prediction models, wherein the final prediction model includes one.

[0159] According to an embodiment of the present disclosure, any multiple modules among the third acquisition module 701, the second determination module 702, the fourth acquisition module and the third determination module can be combined into one module for implementation, or any one of the modules can be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the third acquisition module 701, the second determination module 702, the fourth acquisition module and the third determination module can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation modes of software, hardware and firmware or in an appropriate combination of any of them. Alternatively, at least one of the third acquisition module 701, the second determination module 702, the fourth acquisition module and the third determination module can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding function can be executed.

[0160] Figure 8 The structural block diagram of the prediction model training method and the business processing volume prediction method according to the embodiment of the present disclosure is schematically shown.

[0161] like Figure 8As shown, the electronic device 800 according to an embodiment of the present disclosure includes a processor 801, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 802 or a program loaded from a storage part 808 to a random access memory (RAM) 803. The processor 801 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or a related chipset and / or a dedicated microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 801 may also include an onboard memory for caching purposes. The processor 801 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0162] In RAM 803, various programs and data required for the operation of electronic device 800 are stored. Processor 801, ROM 802 and RAM 803 are connected to each other via bus 804. Processor 801 performs various operations of the method flow according to the embodiment of the present disclosure by executing the program in ROM 802 and / or RAM 803. It should be noted that the program can also be stored in one or more memories other than ROM 802 and RAM 803. Processor 801 can also perform various operations of the method flow according to the embodiment of the present disclosure by executing the program stored in the one or more memories.

[0163] According to an embodiment of the present disclosure, the electronic device 800 may further include an input / output (I / O) interface 805, which is also connected to the bus 804. The electronic device 800 may further include one or more of the following components connected to the I / O interface 805: an input portion 806 including a keyboard, a mouse, etc.; an output portion 807 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage portion 808 including a hard disk, etc.; and a communication portion 809 including a network interface card such as a LAN card, a modem, etc. The communication portion 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to the I / O interface 805 as needed. A removable medium 811, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 810 as needed, so that a computer program read therefrom is installed into the storage portion 808 as needed.

[0164] The present disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or may exist independently without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present disclosure is implemented.

[0165] According to an embodiment of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, it may include but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, an apparatus or a device. For example, according to an embodiment of the present disclosure, the computer-readable storage medium may include the ROM 802 and / or RAM 803 described above and / or one or more memories other than ROM 802 and RAM 803.

[0166] The embodiment of the present disclosure also includes a computer program product, which includes a computer program, and the computer program includes a program code for executing the method shown in the flowchart. When the computer program product is run in a computer system, the program code is used to enable the computer system to implement the training method of the prediction model and the business processing volume prediction method provided by the embodiment of the present disclosure.

[0167] The above functions defined in the system / device of the embodiment of the present disclosure are executed when the computer program is executed by the processor 801. According to the embodiment of the present disclosure, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0168] In one embodiment, the computer program may rely on tangible storage media such as optical storage devices, magnetic storage devices, etc. In another embodiment, the computer program may also be transmitted and distributed in the form of signals on a network medium, and downloaded and installed through the communication part 809, and / or installed from a removable medium 811. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0169] In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 809, and / or installed from the removable medium 811. When the computer program is executed by the processor 801, the above functions defined in the system of the embodiment of the present disclosure are performed. According to the embodiment of the present disclosure, the system, device, means, module, unit, etc. described above can be implemented by a computer program module.

[0170] According to an embodiment of the present disclosure, the program code for executing the computer program provided by the embodiment of the present disclosure can be written in any combination of one or more programming languages. Specifically, these computing programs can be implemented using high-level process and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, Java, C++, python, "C" language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on the remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect through the Internet).

[0171] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0172] It will be appreciated by those skilled in the art that the features described in the various embodiments and / or claims of the present disclosure may be combined and / or combined in a variety of ways, even if such combinations and / or combinations are not explicitly described in the present disclosure. In particular, the features described in the various embodiments and / or claims of the present disclosure may be combined and / or combined in a variety of ways without departing from the spirit and teachings of the present disclosure. All of these combinations and / or combinations fall within the scope of the present disclosure.

[0173] The embodiments of the present disclosure are described above. However, these embodiments are only for illustrative purposes and are not intended to limit the scope of the present disclosure. Although the embodiments are described above separately, this does not mean that the measures in the various embodiments cannot be used in combination to advantage. The scope of the present disclosure is defined by the attached claims and their equivalents. Without departing from the scope of the present disclosure, those skilled in the art may make a variety of substitutions and modifications, which should all fall within the scope of the present disclosure.

Claims

1. A prediction model training method, comprising: Acquire training samples, wherein the training samples include training data and training labels corresponding to the training data, wherein the training data includes data related to handling business, the training labels include historical business processing volume of handling target business at business outlets within a target time period, the training data includes: weather data, date data, time data, geographic location data, and business type data, the date data has at least one of financial attribute information and holiday attribute information, the financial attribute information includes at least one of pension issuance, bond issuance, and commemorative coin issuance, and the time data includes a time interval within a business day; Acquire multiple initial models, wherein the network structures of the multiple initial models are different; Using the training samples to train the multiple initial models to obtain multiple prediction models, wherein each prediction model corresponds to one of the multiple initial models, the prediction model is obtained by adjusting the initial model parameters based on a comparison result, the comparison result is obtained by comparing the training result and the training label, and the training result is determined based on the initial model and the training data; and A target prediction model for predicting the business processing volume is determined from the multiple prediction models.

2. The method according to claim 1, wherein: The initial model includes at least one of the following: Autoregressive moving average model, autoregressive difference moving average model, autoregressive model, moving average model, long short-term memory network model.

3. The method according to claim 1, wherein: Determining a target prediction model from the multiple prediction models comprises: Acquire a test sample, wherein the test sample includes test data and a test tag corresponding to the test data, wherein the test data includes data related to handling a business, and the test tag includes a historical business processing volume of handling a target business within a target time period; Inputting the test data into each of the multiple prediction models to obtain the business processing volume test results of each of the multiple prediction models; and The target prediction model is determined based on the business processing volume test results of each of the multiple prediction models and the test labels.

4. A method for predicting business processing volume, comprising: Obtaining characteristic data related to the business to be predicted within the time period to be predicted; as well as Inputting the characteristic data into the target prediction model to obtain a prediction result of the business processing volume of the business to be predicted within the time period to be predicted; Wherein, the target prediction model is trained using the method according to any one of claims 1 to 3.

5. The method according to claim 4, wherein: The target prediction model includes multiple; The step of inputting the characteristic data into the target prediction model to obtain a prediction result of the business processing volume of the business to be predicted within the time period to be predicted includes: Inputting the characteristic data into a plurality of target prediction models to obtain prediction results of the business processing volume of the business to be predicted within the time period to be predicted for each of the plurality of target prediction models; A target processing volume prediction result is determined based on the business processing volume prediction results of the to-be-predicted business within the to-be-predicted time period of each of the multiple target prediction models.

6. The method according to claim 5, further comprising: Obtaining actual results of the business processing volume of the business to be predicted within the time period to be predicted; as well as Based on the business processing volume forecast results of the multiple target forecast models for the businesses to be forecasted within the forecast time period and the actual business processing volume results, a final forecast model is determined from the multiple target forecast models, wherein the final forecast model includes one.

7. A training device for a prediction model, comprising: A first acquisition module is used to acquire training samples, wherein the training samples include training data and training labels corresponding to the training data, wherein the training data includes data related to handling business, the training labels include historical business processing volume of handling target business at business outlets within a target time period, the training data includes: weather data, date data, time data, geographic location data, and business type data, the date data has at least one of financial attribute information and holiday attribute information, the financial attribute information includes at least one of pension issuance, bond issuance, and commemorative coin issuance, and the time data includes a time interval within a business day; A second acquisition module is used to acquire multiple initial models, wherein the network structures of the multiple initial models are different; a training module, configured to train the plurality of initial models using the training samples to obtain a plurality of prediction models, wherein each of the prediction models corresponds to one of the plurality of initial models, the prediction model is obtained by adjusting the initial model parameters based on a comparison result, the comparison result is obtained by comparing the training result and the training label, and the training result is determined based on the initial model and the training data; and The first determination module is used to determine a target prediction model for predicting the business processing volume from the multiple prediction models.

8. A service processing volume prediction device, comprising: The third acquisition module is used to obtain characteristic data related to the business to be predicted within the time period to be predicted; as well as A second determination module is used to input the characteristic data into a target prediction model to obtain a prediction result of the business processing volume of the business to be predicted within the time period to be predicted; Wherein, the target prediction model is trained using the method according to any one of claims 1 to 3.

9. A computer system comprising: one or more processors; a memory for storing one or more programs, Wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the method of any one of claims 1 to 3 or the method of any one of claims 4 to 6.

10. A computer-readable storage medium having executable instructions stored thereon, wherein when the instructions are executed by a processor, the processor implements the method according to any one of claims 1 to 3 or the method according to any one of claims 4 to 6.

11. A computer program product comprising: Computer executable instructions, which, when executed, are used to implement the method of any one of claims 1 to 3 or the method of any one of claims 4 to 6.

Images