A method, system, electronic device and storage medium for predicting customer flow at a network point

Abstract

The present disclosure provides a method, system, electronic device and storage medium for predicting the passenger flow of outlets, which are applied to the field of financial technology or other fields, including: obtaining the historical passenger flow data and scene description data of each outlet, and generating the historical passenger flow characteristics and scene description characteristics of each outlet based on the data; generating a scene recognition model based on the historical passenger flow characteristics and scene description characteristics; dividing the historical passenger flow data into a training set and a test set, and generating at least one passenger flow prediction model; predicting the passenger flow of the outlet to be predicted based on the scene recognition model and the passenger flow prediction model, and pushing the obtained passenger flow data to the outlet to be predicted. The method of the present disclosure can automatically select different passenger flow prediction models for different bank branches. The present disclosure also provides a system, electronic device and computer-readable storage medium for predicting the passenger flow of outlets.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, and in particular to a method, system, electronic device and storage medium for predicting customer flow at a website. Background Art

[0002] Bank branches are important places for banks to conduct external business. Major bank branches need to receive hundreds of customers every day. In order to effectively provide customers with high-quality services, it is necessary to accurately predict the customer flow of bank branches, reasonably arrange bank branch staff, match bank branch staff with customers, improve the operating efficiency of bank branches, reduce customer waiting time, and improve customer satisfaction.

[0003] Traditional bank branch customer flow forecasting is usually based on the experience of business personnel and simple statistics of historical data, or uses general time series forecasting algorithms, such as exponential smoothing algorithm, ARIMA algorithm, etc. Since the customer flow of each branch is affected by the economic environment, surrounding cultural environment, traffic environment, weather environment, national holidays, major events, etc. of the region, a single time series forecasting algorithm is difficult to apply to all bank branches. Therefore, it is particularly important to automatically select different time series forecasting algorithms for different bank branches. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] In response to the above problems, the present disclosure provides a method, system, electronic device and storage medium for predicting branch customer flow, which are used to at least partially solve technical problems such as the difficulty of applying traditional single time series prediction algorithms to all bank branches.

[0006] (II) Technical solution

[0007] On the one hand, the present disclosure provides a method for predicting the passenger flow of outlets, including: obtaining historical passenger flow data and scene description data of each outlet, and generating historical passenger flow characteristics and scene description characteristics of each outlet based on the data; generating a scene recognition model based on the historical passenger flow characteristics and scene description characteristics; dividing the historical passenger flow data into a training set and a test set, and generating at least one passenger flow prediction model; predicting the passenger flow of the outlet to be predicted based on the scene recognition model and the passenger flow prediction model, and pushing the obtained passenger flow data to the outlet to be predicted.

[0008] Furthermore, generating at least one passenger flow prediction model also includes: obtaining different scene categories according to the scene recognition model, and determining corresponding passenger flow prediction models according to the different scene categories.

[0009] Furthermore, determining the corresponding passenger flow prediction model according to different scene categories includes: calculating the average prediction error of each passenger flow prediction model corresponding to the different scene categories, and determining the passenger flow prediction model corresponding to each scene according to the size of the average prediction error.

[0010] Furthermore, predicting the passenger flow of the outlet to be predicted based on the scene recognition model and the passenger flow prediction model includes: obtaining historical passenger flow data and scene description data of the outlet to be predicted, and generating historical passenger flow characteristics and scene description characteristics of the outlet to be predicted based on this; inputting the historical passenger flow characteristics and scene description characteristics into the scene recognition model to determine the scene category; determining the corresponding passenger flow prediction model according to the scene category, and predicting the passenger flow of the outlet to be predicted based on the passenger flow prediction model.

[0011] Furthermore, determining the corresponding passenger flow prediction model according to the scene category includes: selecting the passenger flow prediction model with the smallest average prediction error corresponding to the scene category according to the scene category, and predicting the passenger flow of the to-be-predicted outlet according to the passenger flow prediction model with the smallest average prediction error.

[0012] Furthermore, the passenger flow prediction model includes a time series prediction algorithm, and the time series prediction algorithm includes a linear regression algorithm, an exponential smoothing algorithm, an ARIMA algorithm, a SARIMA algorithm, a LSTM algorithm, and a random walk algorithm.

[0013] Furthermore, generating historical passenger flow characteristics and scene description characteristics of each outlet based on this includes: generating statistical feature data based on the historical passenger flow data and scene description data of each outlet, including mean, median, standard deviation, 25% quantile, 75% quantile, minimum value and maximum value.

[0014] Furthermore, generating a scene recognition model based on historical passenger flow characteristics and scene description characteristics includes: performing cluster analysis on the historical passenger flow characteristics and scene description characteristics, clustering each network point into different scene categories through a clustering algorithm, and generating a scene recognition model.

[0015] Furthermore, the clustering algorithm includes a hierarchical clustering algorithm.

[0016] On the other hand, the present disclosure provides a network point passenger flow prediction system, including: a first acquisition module, used to acquire historical passenger flow data and scene description data of each network point, and generate historical passenger flow characteristics and scene description characteristics of each network point based on this; a scene recognition module, used to generate a scene recognition model according to the historical passenger flow characteristics and scene description characteristics; a training module, which divides the historical passenger flow data into a training set and a test set, and generates at least one passenger flow prediction model; a push module, which is used to predict the passenger flow of the network point to be predicted according to the scene recognition model and the passenger flow prediction model, and push the predicted passenger flow data to the network point.

[0017] Furthermore, the system also includes: a first determination module, which is used to obtain different scene categories according to the scene recognition model, and determine the corresponding passenger flow prediction model according to the different scene categories.

[0018] Furthermore, the system also includes: a calculation module, which is used to calculate the average prediction error of each passenger flow prediction model corresponding to different scene categories, and determine the passenger flow prediction model corresponding to each scene according to the size of the average prediction error.

[0019] Furthermore, the system also includes: a second acquisition module, used to obtain historical passenger flow data and scene description data of the network point to be predicted, and generate historical passenger flow characteristics and scene description characteristics of the network point to be predicted based on this; a second determination module, used to input the scene recognition model according to the historical passenger flow characteristics and scene description characteristics, and determine the scene category; a prediction module, used to determine the corresponding passenger flow prediction model according to the scene category, and predict the passenger flow of the network point to be predicted based on the passenger flow prediction model.

[0020] Another aspect of the present disclosure provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the processor executes the aforementioned method for predicting customer flow at a branch.

[0021] Another aspect of the present disclosure provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned method for predicting the customer flow of a network point.

[0022] (III) Beneficial effects

[0023] The present invention provides a method, system, electronic device and storage medium for predicting branch customer flow. By dividing bank branches into different scenarios and selecting the best customer flow prediction algorithm for each scenario, the best time series prediction algorithm can be provided for each bank branch, thereby effectively reducing the prediction error of the bank branch customer flow, thereby improving the operating efficiency of the bank branch, reducing customer waiting time, and improving customer satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] For a more complete understanding of the present disclosure and its advantages, reference will now be made to the following description taken in conjunction with the accompanying drawings, in which:

[0025] Figure 1 A schematic diagram schematically shows an application scenario of a method for predicting customer flow at a network point according to an embodiment of the present disclosure;

[0026] Figure 2 A flowchart of a method for predicting customer flow of a network point according to an embodiment of the present disclosure is schematically shown;

[0027] Figure 3 The diagram schematically shows the interaction process between the master node and the sub-node according to the embodiment of the present disclosure;

[0028] Figure 4 A flowchart for predicting the passenger flow of a network point to be predicted according to an embodiment of the present disclosure is schematically shown;

[0029] Figure 5 A flowchart of a method for predicting customer flow of a network point according to a specific embodiment of the present disclosure is schematically shown;

[0030] Figure 6 A block diagram of a passenger flow prediction system according to an embodiment of the present disclosure is schematically shown;

[0031] Figure 7 Another block diagram of a passenger flow prediction system according to an embodiment of the present disclosure is schematically shown;

[0032] Figure 8 A block diagram of an electronic device suitable for implementing the method described above according to an embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0033] 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.

[0034] 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.

[0035] 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.

[0036] 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.).

[0037] Some block diagrams and / or flow charts are shown in the accompanying drawings. It should be understood that some boxes or combinations thereof in the block diagrams and / or flow charts can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that these instructions can create a device for implementing the functions / operations described in these block diagrams and / or flow charts when executed by the processor. The technology of the present disclosure can be implemented in the form of hardware and / or software (including firmware, microcode, etc.). In addition, the technology of the present disclosure can take the form of a computer program product on a computer-readable storage medium storing instructions, which can be used by an instruction execution system or used in conjunction with an instruction execution system.

[0038] The embodiments of the present disclosure provide a method, system, electronic device and storage medium for predicting branch customer flow. Compared with a single time series prediction algorithm, the best time series prediction algorithm can be selected for each bank branch, thereby effectively reducing the prediction error of bank branch customer flow, thereby improving the operating efficiency of bank branches, reducing customer waiting time, and improving customer satisfaction.

[0039] Figure 1 The exemplary system architecture 100 of the method for predicting the customer flow of a network point according to an embodiment of the present disclosure is schematically shown. 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.

[0040] like Figure 1 As shown, the system architecture 100 according to this embodiment may include a terminal device 101, a network 102 and a server 103. The network 102 is used to provide a medium for a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0041] Users (such as developers) can use terminal device 101 to interact with server 103 through network 102 to receive or send messages, etc. Various communication client applications can be installed on terminal device 101, such as various software programming systems, web browser applications, instant messaging tools, email clients, social platform software, etc. (only examples).

[0042] The terminal device 101 may be any electronic device having a display screen and supporting web browsing, including but not limited to a smart phone, a tablet computer, a laptop computer, a desktop computer, and the like.

[0043] The server 103 may be a server that provides various services, such as a background management server (only as an example) that provides support for the application program used by the user using the terminal device 101. The background management server may analyze and process the received user request, etc., and feed back the processing result (such as obtaining a test case file, information or data according to the user request, etc.) to the terminal device.

[0044] It should be noted that the method for predicting the passenger flow of outlets provided in the embodiment of the present disclosure can generally be executed by the server 103. Accordingly, the system for predicting the passenger flow of outlets provided in the embodiment of the present disclosure can generally be deployed in the server 103. The method for predicting the passenger flow of outlets provided in the embodiment of the present disclosure can also be executed by a server or server cluster that is different from the server 103 and can communicate with the terminal device 101 and / or the server 103. Accordingly, the system for predicting the passenger flow of outlets provided in the embodiment of the present disclosure can also be set in a server or server cluster that is different from the server 103 and can communicate with the terminal device 101 and / or the server 103.

[0045] 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 the implementation requirements.

[0046] Figure 2 The flowchart of the method for predicting the customer flow of an outlet according to an embodiment of the present disclosure is schematically shown.

[0047] like Figure 2 As shown in FIG. 1 , the prediction method of the customer flow of the outlet includes:

[0048] In operation S1, historical passenger flow data and scene description data of each outlet are obtained, and historical passenger flow features and scene description features of each outlet are generated based on the data.

[0049] First, the historical passenger flow data of all bank outlets and the description data of bank outlets are obtained; for a single outlet, there are many factors that affect the passenger flow, and the factors that affect the passenger flow data may include time, the working conditions of the outlet, the environment around the outlet, etc., all of which will affect the future passenger flow. In its preferred embodiment, the scene description data of the bank outlet includes but is not limited to: the city level where the outlet is located, the district and county level where the outlet is located, the street (township) level where the outlet is located, the number of permanent residents in the street (township) where the outlet is located, the area of ​​the outlet, the number of windows of the outlet, the number of bus stops within 500 meters around the outlet, and the number of subway stations within 1000 meters around the outlet. The embodiment of the present disclosure generates feature data of bank outlets by acquiring historical passenger flow data and scene description data, and predicts passenger flow based on the feature attributes of these data, which can better improve the accuracy of prediction.

[0050] In operation S2, a scene recognition model is generated according to the historical passenger flow characteristics and the scene description characteristics.

[0051] As an example, the scene recognition model can adopt various unsupervised learning algorithms, such as a hierarchical clustering algorithm.

[0052] In operation S3, the historical passenger flow data is divided into a training set and a test set, and at least one passenger flow prediction model is generated.

[0053] The historical passenger flow data is divided into a training set and a test set; the training set is input into the passenger flow prediction model to train the model; the test set is input into the trained passenger flow prediction model, the prediction result is output, and it is detected whether the prediction result is less than the expected value. If so, the prediction result is output, otherwise the parameters are updated and the test set input step is re-executed.

[0054] In operation S4, the passenger flow of the network point to be predicted is predicted based on the scene recognition model and the passenger flow prediction model, and the obtained passenger flow data is pushed to the network point to be predicted.

[0055] First, the scene category of the bank branch is identified through the bank branch scene recognition model; then the corresponding optimal customer flow prediction model is determined according to the scene category; the future customer flow of the branch to be predicted is predicted according to the optimal customer flow prediction model, and the future customer flow prediction results are sent to the branch to be predicted. Figure 3 The diagram schematically shows the interaction process between the main node and the sub-nodes. The method for predicting the customer flow of the branch usually includes one main node EN and multiple sub-nodes N1-31, N2-32, N3-33, and N4-34 communicating with it. The main node 30 can be a host, which is used to obtain the historical customer flow data and scene description data of each bank branch sub-node, and then predict the customer flow of a single branch to be predicted after obtaining the scene recognition model and the customer flow prediction model, and return the customer flow prediction result to the corresponding sub-node; the sub-nodes 31, 32, 33, and 34 can also be hosts, which are used to reasonably arrange bank branch staff after obtaining the customer flow prediction result forwarded by the main node, realize the matching of bank branch staff and customers, and improve the operating efficiency of bank branches.

[0056] On the basis of the above embodiment, generating at least one passenger flow prediction model further includes: obtaining different scene categories according to the scene recognition model, and determining corresponding passenger flow prediction models according to the different scene categories.

[0057] The output results of the scene recognition model include multiple scene categories. Here, the best passenger flow prediction model is matched according to different scene categories.

[0058] The present invention divides bank branches into different scenarios and selects the best customer flow prediction model for different scenarios. Compared with the traditional single time series prediction algorithm model, it can effectively reduce the prediction error of customer flow of bank branches.

[0059] Based on the above embodiment, determining the corresponding passenger flow prediction model according to different scene categories includes: calculating the average prediction error of each passenger flow prediction model corresponding to different scene categories, and determining the passenger flow prediction model corresponding to each scene according to the size of the average prediction error.

[0060] The best customer flow prediction model is selected by calculating the average prediction error of each customer flow prediction model in each business outlet scenario. The average prediction error is used to predict the average degree to which the output of the prediction model is larger or smaller than the true value. The prediction error includes but is not limited to: Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), etc. The specific calculation method of the average prediction error is not the focus of this disclosure and will not be repeated here.

[0061] Figure 4 A flowchart for predicting the passenger flow of a network point to be predicted according to an embodiment of the present disclosure is schematically shown.

[0062] S501, obtaining historical passenger flow data and scene description data of the network point to be predicted, and generating historical passenger flow features and scene description features of the network point to be predicted based on the data.

[0063] After the scene recognition model and the passenger flow prediction model are determined, the passenger flow data of a single branch to be predicted can be predicted. First, obtain the historical passenger flow data of the bank's business branch and the description data of the bank's business branch. It should be noted that the data of a single branch can be the same as or different from the data of all branches. For example: suppose the data of all branches is from January 1, 2018 to May 1, 2021, and then a scene recognition model and a prediction model are trained based on these data; if the historical data of a single branch is from January 1, 2018 to May 1, 2021, then the span of the historical data of all branches and the single branch is the same; if the historical data of a single branch is from January 1, 2019 to May 1, 2021, then the span of the historical data of all branches and the single branch is different, and the data is also different.

[0064] S502, inputting historical passenger flow characteristics and scene description characteristics into a scene recognition model to determine the scene category.

[0065] The corresponding scene category is obtained according to the historical customer flow characteristics and scene description characteristics of the outlets to be predicted. Specifically, each bank outlet can be divided into categories such as A, B, C, D, etc. according to the historical customer flow characteristics and scene description characteristics of the outlets. Each category of outlets has the same or similar feature data.

[0066] S503, determining a corresponding passenger flow prediction model according to the scene category, and predicting the passenger flow of the network point to be predicted according to the passenger flow prediction model.

[0067] After determining the scenario category, the corresponding customer flow prediction model is determined. The customer flow prediction model is the optimal prediction model under the scenario. Predicting the customer flow of the predicted outlets is conducive to improving the operational efficiency of bank outlets and improving customer satisfaction.

[0068] Based on the above embodiments, determining the corresponding passenger flow prediction model according to the scene category includes: according to the scene category, selecting the passenger flow prediction model with the smallest average prediction error corresponding to the scene category, and predicting the passenger flow of the outlet to be predicted based on the passenger flow prediction model with the smallest average prediction error.

[0069] The average prediction error of each customer flow prediction model in each business outlet scenario is calculated, and the customer flow prediction model with the smallest error is finally determined as the prediction model corresponding to the scenario. Thus, after determining the scenario category, the corresponding customer flow prediction model can be determined, and the customer flow prediction model is used to predict the customer flow of the outlet to be predicted, which is conducive to obtaining more accurate customer flow prediction results.

[0070] Based on the above embodiment, the passenger flow prediction model includes a time series prediction algorithm, and the time series prediction algorithm includes a linear regression algorithm, an exponential smoothing algorithm, an ARIMA algorithm, a SARIMA algorithm, a LSTM algorithm, and a random walk algorithm.

[0071] The basic idea of ​​the time series data algorithm is: for a large amount of time series data, according to the purpose of the application, select the appropriate mining method, and then mine the implicit rules from the existing time series data, and then use these rules to predict or describe the future trend of the time series data. Among them, the ARIMA algorithm model is very flexible and combines the advantages of time series analysis and regression analysis. The occurrence of passenger flow is a dynamic and complex system with the characteristics of unevenness, difference, diversity, suddenness, randomness, predictability and regularity. A single prediction model is based on linear or nonlinear data and cannot capture the linear and nonlinear laws of its occurrence at the same time, so it is difficult to achieve ideal prediction accuracy. The differential autoregressive moving average model (ARIMA) is used to predict the historical data of passenger flow, so that its linear law information is included in the prediction results of the ARIMA model. At this time, the nonlinear law is included in the residual sequence, which is conducive to improving the accuracy of the predicted passenger flow.

[0072] Based on the above embodiments, generating historical passenger flow characteristics and scene description characteristics of each outlet includes: generating statistical feature data based on the historical passenger flow data and scene description data of each outlet, including mean, median, standard deviation, 25% quantile, 75% quantile, minimum value and maximum value.

[0073] The historical customer flow of the outlet, the geographical location of the outlet, and the size of the outlet also affect the future customer flow trend of the outlet. This type of data belongs to the historical statistical characteristics and scene description characteristics of the outlet. The historical statistical feature data can be taken as the mean, median, etc., and the specific method can be determined according to the actual situation.

[0074] Based on the above embodiments, generating a scene recognition model according to historical passenger flow characteristics and scene description characteristics includes: clustering the historical passenger flow characteristics and scene description characteristics, clustering each network point into different scene categories through a clustering algorithm, and generating a scene recognition model.

[0075] The historical passenger flow characteristics and scene description characteristics are used as the data basis for clustering. The clustering algorithm is used to cluster each outlet into different scene categories, so that the scene category of the outlet to be predicted is determined according to the scene recognition model.

[0076] Based on the above embodiment, the clustering algorithm includes a hierarchical clustering algorithm.

[0077] As an unsupervised learning method, cluster analysis algorithm is widely used in data processing in various industries, such as hierarchical clustering algorithm, K-means algorithm, etc.

[0078] The method of the disclosed embodiment selects the best time series prediction algorithm for different scenarios, effectively reducing the prediction error of customer flow at bank branches, thereby improving the operating efficiency of bank branches, reducing customer waiting time, and improving customer satisfaction.

[0079] Combine the following Figure 5 The specific implementation instructions for each step of this method are given.

[0080] like Figure 5 As shown, the present disclosure provides a method for predicting the customer flow of a bank branch, comprising the following steps:

[0081] Step 1: Obtain the historical customer flow data of all bank branches and the description data of bank branches;

[0082] In its preferred embodiment, the descriptive data of the bank branch offices include but are not limited to: the city level where the branch office is located, the district or county level where the branch office is located, the street (township) level where the branch office is located, the number of permanent residents in the street (township) where the branch office is located, the area of ​​the branch office, the number of windows of the branch office, the number of bus stops within 500 meters around the branch office, and the number of subway stations within 1000 meters around the branch office.

[0083] Step 2: Generate characteristic data of bank branches based on the description data and historical customer flow data of bank branches;

[0084] In a preferred embodiment, statistical features are generated based on historical passenger flow data, including but not limited to: mean, median, standard deviation, 25% quantile, 75% quantile, minimum and maximum values, etc.; which is equivalent to the aforementioned step S1.

[0085] Step 3: Perform cluster analysis on the characteristic data of all business outlets to generate a scene recognition model for bank business outlets;

[0086] In a preferred embodiment, the clustering algorithm adopts a hierarchical clustering algorithm.

[0087] In a preferred embodiment, bank branches are clustered into different scenario categories through a clustering algorithm, which is equivalent to the aforementioned step S2.

[0088] Step 4: Divide the historical customer flow data of bank branches into training sets and test sets, and use the training sets and test sets to evaluate the time series prediction algorithm in each scenario to determine the best time series prediction model corresponding to the scenario.

[0089] In its preferred embodiment, the time series prediction algorithm includes but is not limited to: linear regression algorithm, exponential smoothing algorithm, ARIMA algorithm, SARIMA algorithm, LATM algorithm, random walk algorithm, etc.

[0090] In a preferred embodiment, the historical time series of each bank branch is split into a training set and a test set.

[0091] In a preferred embodiment, for each business outlet, a time series prediction algorithm is trained using a training set, and the customer flow at a corresponding time point of a test set is predicted.

[0092] In a preferred embodiment thereof, the best time series prediction algorithm is selected by calculating the average prediction error of each prediction algorithm in each business outlet scenario.

[0093] In a preferred embodiment, the prediction error includes but is not limited to: root mean square error, mean absolute error, mean absolute percentage error, etc.; which is equivalent to the aforementioned step S3.

[0094] Step 5: For any bank branch’s historical customer flow data, bank branch’s description data, bank branch’s scene recognition model and bank branch’s time series prediction model, predict the bank branch’s future customer flow.

[0095] 5.1. For each bank branch, obtain the historical customer flow data and description data of the bank branch;

[0096] 5.2. Generate characteristic data of bank branches using description data and historical customer flow data of bank branches;

[0097] 5.3. Use the bank branch scene recognition model generated in the third step to identify the scene category of the bank branch;

[0098] 5.4. Use the best time series prediction model corresponding to each scenario generated in step 4 to determine the prediction algorithm model of the bank branch;

[0099] 5.5. Based on the historical customer flow data and the prediction algorithm model, the future customer flow of the bank's business outlets is predicted; this is equivalent to the aforementioned step S4.

[0100] It should be noted that a single time series prediction algorithm is used by all outlets; while the present application uses different prediction algorithms for different outlets. Assume that there are 2,000 outlets, and the single time prediction algorithm uses algorithm A; in the present application, there may be 500 outlets using algorithm A, 500 outlets using algorithm B, and so on. In addition, in steps 1 to 4, two models are produced using all business outlets: a scene recognition model and a prediction model for each scene; then in step 5, for any business outlet, the scene category of the business outlet is first identified using the scene recognition model, the corresponding prediction model is selected according to the category, and then the prediction model is used to predict the future passenger flow of the business outlet.

[0101] Figure 6 A block diagram of a network point passenger flow prediction system according to an embodiment of the present disclosure is schematically shown.

[0102] like Figure 6 As shown, the passenger flow prediction system 600 includes: a first acquisition module 610, a scene recognition module 620, a training module 630, and a push module 640.

[0103] The first acquisition module 610 is used to acquire the historical passenger flow data and scene description data of each network point, and generate the historical passenger flow characteristics and scene description characteristics of each network point based on the data; according to the embodiment of the present disclosure, the first acquisition module 610 can be used to execute the above reference Figure 2 The S1 step described is not repeated here.

[0104] The scene recognition module 620 is used to generate a scene recognition model according to the historical passenger flow characteristics and the scene description characteristics; according to the embodiment of the present disclosure, the scene recognition module 620 can be used to perform the above reference Figure 2 The S2 step is described.

[0105] The training module 630 divides the historical passenger flow data into a training set and a test set to generate at least one passenger flow prediction model; according to an embodiment of the present disclosure, the training module 630 can be used to perform the above reference Figure 2 The S3 step described.

[0106] The push module 640 is used to predict the passenger flow of the predicted outlets according to the scene recognition model and the passenger flow prediction model, and push the predicted passenger flow data to the outlets. According to an embodiment of the present disclosure, the push module 640 can be used to execute the above reference Figure 2 Step S4 as described.

[0107] like Figure 7 As shown, the network point passenger flow prediction system 600 also includes: a first determination module 650, a calculation module 660, a second acquisition module 670, a second determination module 680, and a prediction module 690.

[0108] The first determination module 650 is used to obtain different scene categories according to the scene recognition model, and determine the corresponding passenger flow prediction model according to the different scene categories; according to an embodiment of the present disclosure, the first determination module 650 can be used to execute the steps of determining the corresponding passenger flow prediction model according to the different scene categories described above, which will not be repeated here.

[0109] The calculation module 660 is used to calculate the average prediction error of each passenger flow prediction model corresponding to different scene categories, and determine the passenger flow prediction model corresponding to each scene according to the size of the average prediction error. According to an embodiment of the present disclosure, the calculation module 660 can be used to perform the steps of calculating the average prediction error of each passenger flow prediction model described above, which will not be repeated here.

[0110] The second acquisition module 670 is used to acquire the historical passenger flow data and scene description data of the network point to be predicted, and generate the historical passenger flow characteristics and scene description characteristics of the network point to be predicted. According to the embodiment of the present disclosure, the second acquisition module 670 can be used to perform the above reference Figure 4 The S501 step described.

[0111] The second determination module 680 is used to input the scene recognition model according to the historical passenger flow characteristics and the scene description characteristics to determine the scene category. According to an embodiment of the present disclosure, the second determination module 680 can be used to perform the above reference Figure 4 The described step S502.

[0112] The prediction module 690 is used to determine the corresponding passenger flow prediction model according to the scene category, and predict the passenger flow of the predicted network point according to the passenger flow prediction model. According to the embodiment of the present disclosure, the prediction module 690 can be used to perform the above reference Figure 4 The described step S503.

[0113] It should be noted that any one or more of the modules, submodules, units, and subunits according to the embodiments of the present disclosure, or at least part of the functions of any one of them, can be implemented in one module. Any one or more of the modules, submodules, units, and subunits according to the embodiments of the present disclosure can be split into multiple modules for implementation. Any one or more of the modules, submodules, units, and subunits according to the embodiments of the present disclosure 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 in any other reasonable way of integrating or packaging the circuit, or in any one of the three implementation methods of software, hardware, and firmware, or in any appropriate combination of any of them. Alternatively, one or more of the modules, submodules, units, and subunits according to the embodiments of the present disclosure can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding function can be performed.

[0114] For example, any multiple of the first acquisition module 610, the scene recognition module 620, the training module 630, the push module 640, the first determination module 650, the calculation module 660, the second acquisition module 670, the second determination module 680, and the prediction module 690 may be combined into one module for implementation, or any one of the modules may be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules may 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 610, the scene recognition module 620, the training module 630, the push module 640, the first determination module 650, the calculation module 660, the second acquisition module 670, the second determination module 680, and the prediction module 690 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 by any one of the three implementation methods of software, hardware and firmware or by an appropriate combination of any of them. Alternatively, at least one of the first acquisition module 610, the scene recognition module 620, the training module 630, the push module 640, the first determination module 650, the calculation module 660, the second acquisition module 670, the second determination module 680, and the prediction module 690 can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding function can be performed.

[0115] The branch passenger flow prediction system and method provided by the present disclosure can be used in the financial field or other fields. It should be noted that the branch passenger flow prediction system and method provided by the present disclosure can be used in the financial field, such as the passenger flow prediction of various business branches of banks in the financial field, and can also be used in any field except the financial field. The application field of the branch passenger flow prediction system and method provided by the present disclosure is not limited.

[0116] Figure 8 A block diagram of an electronic device suitable for implementing the method described above according to an embodiment of the present disclosure is schematically shown. Figure 8 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0117] like Figure 8As shown, the electronic device 800 described in this embodiment 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 (such as a CPU), an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (for example, 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.

[0118] In RAM 803, various programs and data required for the operation of system 800 are stored. Processor 801, ROM 802 and RAM 803 are connected to each other through 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 one or more memories.

[0119] 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 system 800 may further include one or more of the following components connected to the I / O interface 805: an input section 806 including a keyboard, a mouse, etc.; an output section 807 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 808 including a hard disk, etc.; and a communication section 809 including a network interface card such as a LAN card, a modem, etc. The communication section 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 section 808 as needed.

[0120] According to an embodiment of the present disclosure, the method flow according to an embodiment of the present disclosure can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable storage medium, and the computer program contains a program code for executing the method shown in the flowchart. 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-mentioned functions defined in the system of the embodiment of the present disclosure are executed. According to an embodiment of the present disclosure, the system, equipment, device, module, unit, etc. described above can be implemented by a computer program module.

[0121] The embodiments of the present disclosure also provide 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 for predicting the customer flow of a network point according to the embodiments of the present disclosure is implemented.

[0122] According to an embodiment of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, 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 of the above. In an embodiment of the present disclosure, the 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.

[0123] It should be noted that the functional modules in the various embodiments of the present disclosure may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated modules may be implemented in the form of hardware or in the form of software functional modules. If the integrated modules are implemented in the form of software functional modules and sold or used as independent products, they may be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present disclosure may essentially or in other words, the part that contributes to the prior art or all or part of the technical solution may be embodied in the form of a software product.

[0124] 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.

[0125] 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.

[0126] Although the present disclosure has been shown and described with reference to specific exemplary embodiments of the present disclosure, it should be understood by those skilled in the art that various changes in form and details may be made to the present disclosure without departing from the spirit and scope of the present disclosure as defined by the appended claims and their equivalents. Therefore, the scope of the present disclosure should not be limited to the above-mentioned embodiments, but should be determined not only by the appended claims, but also by the equivalents of the appended claims.

Claims

1. A method for predicting customer flow at a network point, characterized in that: include: Obtain historical passenger flow data and scene description data of each outlet corresponding to multiple sub-nodes according to the main node, and generate historical passenger flow characteristics and scene description characteristics of each outlet accordingly; Generate a scene recognition model according to the historical passenger flow characteristics and scene description characteristics; Dividing the historical passenger flow data into a training set and a test set, and generating at least one passenger flow prediction model, including: Obtaining different scene categories according to the scene recognition model; Calculate the average prediction error of each passenger flow prediction model corresponding to different scene categories, wherein the average prediction error is used to predict the average degree to which the output result of the passenger flow prediction model is larger or smaller than the true value; The passenger flow of the predicted network point is predicted according to the scene recognition model and the passenger flow prediction model, and the obtained passenger flow data is pushed to the predicted network points corresponding to the plurality of child nodes according to the master node, wherein the passenger flow of the predicted network point is predicted according to the scene recognition model and the passenger flow prediction model, comprising: Acquire the historical passenger flow data and scene description data of the network point to be predicted, and generate the historical passenger flow characteristics and scene description characteristics of the network point to be predicted based on the data; Inputting the historical passenger flow characteristics and scene description characteristics into a scene recognition model to determine the scene category; According to the scenario category, a passenger flow prediction model with the smallest average prediction error corresponding to the scenario category is selected, and the passenger flow of the outlet to be predicted is predicted based on the passenger flow prediction model with the smallest average prediction error.

2. The method for predicting the customer flow of a network point according to claim 1, characterized in that: The passenger flow prediction model includes a time series prediction algorithm, and the time series prediction algorithm includes a linear regression algorithm, an exponential smoothing algorithm, an ARIMA algorithm, a SARIMA algorithm, a LSTM algorithm, and a random walk algorithm.

3. The method for predicting the customer flow of a network point according to claim 1, characterized in that: The historical passenger flow characteristics and scene description characteristics of each network point generated thereby include: Statistical feature data is generated based on the historical passenger flow data and scene description data of each outlet, including mean, median, standard deviation, 25% quantile, 75% quantile, minimum value and maximum value.

4. The method for predicting the customer flow of a network point according to claim 3, characterized in that: Generating a scene recognition model according to the historical passenger flow characteristics and scene description characteristics includes: The historical passenger flow characteristics and scene description characteristics are clustered and analyzed, and each network point is clustered into different scene categories through a clustering algorithm to generate a scene recognition model.

5. The method for predicting the customer flow of a network point according to claim 4, characterized in that: The clustering algorithm includes a hierarchical clustering algorithm.

6. A network point passenger flow prediction system, characterized in that: include: A first acquisition module is used to acquire historical passenger flow data and scene description data of each network point corresponding to multiple sub-nodes according to the main node, and generate historical passenger flow characteristics and scene description characteristics of each network point accordingly; A scene recognition module, used to generate a scene recognition model according to the historical passenger flow characteristics and scene description characteristics; A training module, dividing the historical passenger flow data into a training set and a test set, and generating at least one passenger flow prediction model; A first determination module, used to obtain different scene categories according to the scene recognition model, and determine corresponding passenger flow prediction models according to the different scene categories; A calculation module, used to calculate the average prediction error of each passenger flow prediction model corresponding to different scene categories, wherein the average prediction error is used to predict the average degree to which the output result of the passenger flow prediction model is larger or smaller than the true value; A push module is used to predict the passenger flow of the predicted outlets according to the scene recognition model and the passenger flow prediction model, and push the predicted passenger flow data to the outlets corresponding to the multiple child nodes according to the master node. The push module includes: A second acquisition module is used to acquire the historical passenger flow data and scene description data of the network point to be predicted, and generate the historical passenger flow characteristics and scene description characteristics of the network point to be predicted based on the historical passenger flow data and scene description characteristics of the network point to be predicted; A second determination module is used to input a scene recognition model according to the historical passenger flow characteristics and scene description characteristics to determine the scene category; The prediction module is used to select the passenger flow prediction model with the smallest average prediction error corresponding to the scene category according to the scene category, and predict the passenger flow of the network point to be predicted according to the passenger flow prediction model with the smallest average prediction error.

7. An electronic device comprising: processor; A memory storing a computer executable program, which, when executed by the processor, enables the processor to execute the method for predicting the customer flow of a network point as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, it implements the method for predicting the customer flow of a network point as described in any one of claims 1 to 5.

Images