An account opening review method and related devices based on LightGBM

Abstract

The present invention provides an account opening review method and related devices based on LightGBM, which can be applied to the financial field or other fields, including obtaining account opening material information collected by bank customer managers; predicting the account opening material information based on a pre-constructed data review model to obtain a prediction result; if the prediction result is qualified, submitting the account opening material information to the bank system to open the corresponding corporate account. In this solution, there is no need for manual review. The account opening material information is used as the input of the pre-constructed data review model, and the data review model constructed based on the LightGBM algorithm model is used to predict the account opening material information to obtain a prediction result to determine whether the account opening material information is qualified; when it is determined that the prediction result is qualified, the account opening material information is submitted to the bank system to open the corresponding corporate account. Through the above method, the processing speed of opening corporate accounts can be improved, which is relatively slow.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular, to an account opening review method based on LightGBM and related devices. Background Art

[0002] In order to prevent shell companies and telecom fraud, currently when opening a corporate account, banks need to review the account opening material information collected by bank account managers on-site from customers.

[0003] Currently, after the bank account manager uploads the on-site collected account opening material information of the customer to the bank system, bank personnel will conduct the review; when the bank personnel find that the account opening material information is missing, they will notify the bank account manager to go to the customer site again to re-collect, and then re-upload the changed account opening material information. Due to the need to wait for manual review, the bank account manager cannot receive feedback immediately at the customer site and may need to re-collect at the customer site. As a result, the processing speed of opening a corporate account is relatively slow. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide an account opening review method based on LightGBM and related devices to solve the problem of relatively slow processing speed of opening a corporate account in the prior art.

[0005] To achieve the above object, embodiments of the present invention provide the following technical solutions:

[0006] A first aspect of an embodiment of the present invention shows an account opening review method based on LightGBM, the method includes:

[0007] Obtain the account opening material information collected by the bank account manager, where the account opening material information at least includes the company legal person information, business address information, and employee situation information collected on-site;

[0008] Use the account opening material information as the input of a pre-constructed data review model, and based on the pre-constructed data review model, predict the account opening material information to obtain a prediction result, where the prediction result is used to indicate whether the account opening material information is qualified, and the pre-constructed data review model is constructed based on the LightGBM algorithm model;

[0009] If the prediction result is qualified, submit the account opening material information to the bank system for opening the corresponding corporate account.

[0010] Optionally, it further includes:

[0011] If the prediction result is unqualified, output a prompt message to prompt the bank account manager to re-collect the account opening material information.

[0012] Optionally, the pre-construction process of the data review model includes:

[0013] Obtain sample data, where the sample data includes historical account opening material information that has passed the review within a historical time period;

[0014] Process the sample data to extract the sample feature data;

[0015] Determine an initial LightGBM algorithm model, and train the initial LightGBM algorithm model based on the sample feature data;

[0016] Evaluate the trained LightGBM algorithm model until the obtained evaluation result is within a preset range, and determine the currently trained LightGBM algorithm model as the data review model.

[0017] Optionally, the training of the initial LightGBM algorithm model based on the sample feature data includes:

[0018] Based on the LightGBM algorithm, configure the initial network parameters of the initial LightGBM algorithm model;

[0019] Divide the sample feature data into a training set, a validation set, and a test set;

[0020] Use the training set to train the initial network parameters, and adjust the initial network parameters based on the validation set to obtain adjusted network parameters;

[0021] Construct a LightGBM algorithm model based on the adjusted network parameters.

[0022] Optionally, the evaluation of the trained LightGBM algorithm model until the obtained evaluation result is within a preset range, and determining the currently trained LightGBM algorithm model as the data review model includes:

[0023] Use the test set to evaluate the LightGBM algorithm model to obtain an evaluation result;

[0024] If the evaluation result is within the preset range, determine the current LightGBM algorithm model as the data review model;

[0025] If the evaluation result is not within the predicted range, continue to train the current LightGBM algorithm model based on the evaluation result, the training set, the validation set, and the test set.

[0026] The second aspect of the embodiments of the present invention shows an account opening review device based on LightGBM, and the device includes:

[0027] An acquisition unit for acquiring the account opening material information collected by a bank customer manager, where the account opening material information at least includes the company legal person information, business address information, and employee situation information collected on-site;

[0028] A data review model for using the account opening material information as an input to a pre-constructed data review model, predicting the account opening material information based on the pre-constructed data review model to obtain a prediction result, where the prediction result is used to indicate whether the account opening material information is qualified, and the pre-constructed data review model is constructed based on a construction unit. If the prediction result is qualified, the submission unit is executed;

[0029] The submission unit for submitting the account opening material information to the bank system to open a corresponding corporate account.

[0030] Optionally, it further includes:

[0031] An output unit for outputting a prompt message when it is determined that the prediction result is unqualified to prompt the bank customer manager to re-collect the account opening material information.

[0032] Optionally, the construction unit includes:

[0033] An acquisition module for acquiring sample data, where the sample data includes the historical account opening material information that has passed the review within a historical time period;

[0034] An extraction module for processing the sample data and extracting the sample feature data;

[0035] A training module for determining an initial LightGBM algorithm model and training the initial LightGBM algorithm model based on the sample feature data;

[0036] An evaluation module for evaluating the trained LightGBM algorithm model until the obtained evaluation result is within a preset range, and determining the currently trained LightGBM algorithm model as the data review model.

[0037] A third aspect of the embodiments of the present invention shows an electronic device for running a program, where the program, when running, executes the account opening review method based on LightGBM shown in the first aspect of the embodiments of the present invention.

[0038] A fourth aspect of the embodiments of the present invention shows a computer storage medium, where the storage medium includes a stored program, and when the program runs, it controls the device where the storage medium is located to execute the account opening review method based on LightGBM shown in the first aspect of the embodiments of the present invention.

[0039] Based on the account opening review method and related devices provided by the embodiments of the present invention described above, the method includes: obtaining the account opening material information collected by the bank customer manager, where the account opening material information at least includes the company legal person information, business address information, and employee situation information collected on-site; using the account opening material information as the input of a pre-constructed data review model, and predicting the account opening material information based on the pre-constructed data review model to obtain a prediction result, where the prediction result is used to indicate whether the account opening material information is qualified, and the pre-constructed data review model is constructed based on the LightGBM algorithm model; if the prediction result is qualified, submitting the account opening material information to the bank system for opening the corresponding corporate account. In the embodiments of the present invention, there is no need for manual review. The account opening material information is used as the input of a pre-constructed data review model, and the data review model constructed based on the LightGBM algorithm model is used to predict the account opening material information to obtain a prediction result to determine whether the account opening material information is qualified; when it is determined that the prediction result is qualified, the account opening material information is submitted to the bank system for opening the corresponding corporate account. Through the above method, the processing speed of opening corporate accounts can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on the provided drawings.

[0041] Figure 1 It is a schematic flowchart of an account opening review method based on LightGBM shown in the embodiments of the present invention;

[0042] Figure 2 It is a schematic flowchart of constructing a data review model based on the LightGBM algorithm model shown in the embodiments of the present invention;

[0043] Figure 3 It is a flowchart of the construction and use of the data review model shown in the embodiments of the present invention;

[0044] Figure 4 It is a schematic structural diagram of an account opening review device based on LightGBM shown in the embodiments of the present invention;

[0045] Figure 5 It is a schematic structural diagram of another account opening review device based on LightGBM shown in the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0047] In this application, the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements but also includes other elements not expressly listed, or also includes elements inherent to such a process, method, article or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article or device including the said element.

[0048] It should be noted that a method and related device for account opening review based on LightGBM provided by the present invention can be used in the fields of artificial intelligence, blockchain, distributed, cloud computing, big data, Internet of Things, mobile Internet, network security, chip, virtual reality, augmented reality, holographic technology, quantum computing, quantum communication, quantum measurement, digital twin or finance. The above are only examples and do not limit the application fields of a method and related device for account opening review based on LightGBM provided by the present invention.

[0049] In the embodiments of the present invention, manual review is not required. The account opening material information is used as the input of a pre-constructed data review model, and the data review model constructed based on the LightGBM algorithm model is used to predict the account opening material information to obtain a prediction result, so as to determine whether the account opening material information is qualified; when it is determined that the prediction result is qualified, the account opening material information is submitted to the bank system to open the corresponding corporate account. Through the above method, the processing speed of opening a corporate account can be improved.

[0050] See Figure 1 , which is a schematic flowchart of a method for account opening review based on LightGBM shown in the embodiments of the present invention. The method includes:

[0051] S101: Obtain the account opening material information collected by the bank customer manager.

[0052] In step S101, the account opening material information at least includes the company legal person information, business address information and employee situation information collected on site.

[0053] In an embodiment of the present invention, in order to prevent shell companies and telecommunications fraud, when opening a corporate account, the bank customer manager of a corporate customer currently needs to collect on-site the information of the customer's account opening materials. In the process of specifically implementing step S101, the information of the company legal person, business address information, employee information, etc., which are collected on-site and uploaded by the bank customer manager of the corporate customer, is obtained.

[0054] It should be noted that the account opening material information is not limited to the company legal person information, business address information, and employee information collected on-site shown above, but also includes information such as the storefront and factory building conditions.

[0055] S102: Use the account opening material information as the input of a pre-constructed data review model, and predict the account opening material information based on the pre-constructed data review model to obtain a prediction result.

[0056] In step S102, the prediction result is used to indicate whether the account opening material information is qualified, and the pre-constructed data review model is constructed based on the LightGBM algorithm model.

[0057] In an embodiment of the present invention, LightGBM is a machine learning algorithm, an open-source framework implemented by Microsoft for the Gradient Boosting Decision Tree (GBDT) algorithm, which supports highly efficient parallel training. The main reason for the proposal of LightGBM is to solve the problems encountered by GBDT in dealing with massive data, enabling GBDT to be better and faster applied to industrial practice, and being famous for its faster training speed, lower memory consumption, and better accuracy, and is commonly used in the financial field.

[0058] It should be noted that in the process of constructing a data review model based on the LightGBM algorithm model, as Figure 2 shown.

[0059] In an embodiment of the present application, based on Figure 2 the schematic flowchart of constructing a data review model based on the LightGBM algorithm model shown, correspondingly, an embodiment of the present application also discloses a flowchart of the construction and use of the data review model, as Figure 3 shown.

[0060] The process of constructing a data review model based on the LightGBM algorithm model includes the following steps:

[0061] S201: Obtain sample data.

[0062] In step S201, the sample data includes historical account opening material information that has passed the review during a historical time period.

[0063] In the process of specifically implementing step S201, obtain historical account opening material information such as the information of corporate legal persons, business address information, and employee situation information that have passed the review within a historical time period, and establish a verification data wide table.

[0064] It should be noted that the historical time period is preset by technical personnel, for example, it can be set as the past year.

[0065] S202: Process the sample data and extract the sample feature data.

[0066] In the process of specifically implementing step S202, perform feature engineering on the sample data to obtain numerical sample feature data.

[0067] It can be understood that the specific process of performing feature engineering on the sample data includes feature selection, such as encoding the legal person information and feature changes, such as encoding the business address, feature derivation, feature enhancement, and other operations.

[0068] S203: Determine the initial LightGBM algorithm model and train the initial LightGBM algorithm model based on the sample feature data.

[0069] In the process of specifically implementing step S203, use the LightGBM model to model the sample feature data after feature engineering. For example: Assume that X is the data after on-site verification and feature engineering, that is, the sample feature data; Y is the review result (review result 0 indicates qualified review; review result 1 indicates unqualified review), input X and Y into the LightGBM model, and use the LightGBM algorithm to model it.

[0070] In the embodiment of the present invention, since LightGBM has the advantages of fast modeling, low memory occupancy, and high accuracy, and is suitable for processing various types of data, the LightGBM algorithm model is used in the present invention to establish a data review model.

[0071] It should be noted that in the process of specifically implementing step S203, the following steps are included:

[0072] S11: Configure the initial network parameters of the initial LightGBM algorithm model based on the LightGBM algorithm.

[0073] S12: Divide the sample feature data into a training set, a validation set, and a test set.

[0074] S13: Use the training set to train the initial network parameters and adjust the initial network parameters based on the validation set to obtain the adjusted network parameters.

[0075] In the process of specifically implementing step S13, the k-fold cross-validation method is adopted to optimize the network parameters trained by the training set on the validation set, obtain the adjusted network parameters, and construct an optimized LightGBM algorithm model based on the adjusted network parameters. Specifically, based on the sample data in the validation set, the network parameters trained by the training set are divided into k parts by the k-fold cross-validation method. The network parameters trained by 1 part of the training set are used as validation data, and the network parameters trained by the other k - 1 parts of the training set are used as training data. The LightGBM algorithm model is trained based on the training data in the validation set, and the validation data is used to predict by the LightGBM algorithm model trained by the training data to obtain the first prediction data.

[0076] For each network parameter trained by the training set, it needs to be verified once. The average value of the k prediction data is taken, that is, the adjusted network parameters, so as to reduce the generalization error, and then step S14 is executed.

[0077] Furthermore, it should be noted that in order to ensure the stability of the model, the k-fold cross-validation method shown in the present invention is the 5-fold cross-validation method, that is to say, k can be 5.

[0078] S14: Construct a LightGBM algorithm model based on the adjusted network parameters.

[0079] In the process of specifically implementing step S14, a LightGBM algorithm model is constructed by the adjusted network parameters.

[0080] S204: Evaluate the trained LightGBM algorithm model until the obtained evaluation result is within the preset range, and determine that the currently trained LightGBM algorithm model is the data review model.

[0081] Optionally, based on the pre-construction process of the model shown above, in the process of executing step S204 to evaluate the trained LightGBM algorithm model until the obtained evaluation result is within the preset range and determining that the currently trained LightGBM algorithm model is the data review model, the following steps are included:

[0082] S21: Use the test set to evaluate the LightGBM algorithm model to obtain an evaluation result.

[0083] In the process of specifically implementing step S21, the test set is predicted in the LightGBM algorithm model to evaluate the LightGBM algorithm model and obtain the corresponding evaluation result.

[0084] S22: Determine whether the evaluation result is within a preset range. If the evaluation result is within the preset range, execute step S23; if the evaluation result is not within the predicted range, execute step S14.

[0085] It should be noted that the preset range refers to the value set by technicians in advance to determine whether the prediction result is qualified.

[0086] S23: Determine the current LightGBM algorithm model as the data review model.

[0087] S24: Continue to train the current LightGBM algorithm model based on the evaluation result, training set, validation set, and test set, that is, return to execute step S202.

[0088] Based on the data review model constructed in the above embodiments, during the specific implementation of step S102, when the bank customer manager verifies information on-site, the account opening material information is input into Figure 3 the constructed data review model shown, so that the constructed data review model predicts the input account opening material information to obtain a prediction result. Among them, the prediction result is a probability value.

[0089] It should be noted that the constructed data review model can be set in many general or special portable computing device environments or configurations. For example: portable personal computers, portable handheld devices or portable devices, portable tablet devices, and distributed computing environments including any of the above devices or equipment, etc.

[0090] S103: Determine whether the prediction result is qualified or unqualified. If the prediction result is qualified, execute step S104; if the prediction result is unqualified, execute step S105.

[0091] During the specific implementation of step S103, determine whether the prediction result is greater than or equal to the preset probability. If it is determined that the prediction result is greater than or equal to the preset probability, execute step S104; if it is determined that the prediction result is less than the preset probability, execute step S105.

[0092] It should be noted that the preset probability is set by technicians in advance based on experience or according to the actual situation.

[0093] S104: Submit the account opening material information that has passed the review to the bank system for opening the corresponding corporate account.

[0094] During the specific implementation of step S104, submit the account opening material information that has passed the review to the bank system for the bank to open the corresponding corporate account.

[0095] S105: Output prompt information to prompt the bank account manager to re-collect account opening material information.

[0096] In the specific implementation of step S105, for data with a low passing probability, a prompt message is output to prompt the bank account manager to re-collect the account opening materials, that is, to supplement the data in real time on site, thereby eliminating the need to wait for the results of manual review and improving the work efficiency of the account manager.

[0097] It should be noted that after executing step S105, the bank account manager re-acquires the account opening material information, that is, returns to execute step S101.

[0098] In the embodiment of the present invention, manual review is not required. The account opening material information is used as the input of the pre-built data review model. The data review model based on the LightGBM algorithm model is used to predict the account opening material information to obtain the prediction result to determine whether the account opening material information is qualified; when the prediction result is determined to be qualified, the account opening material information is submitted to the bank system to open the corresponding public account. The above method can improve the processing speed of opening a public account.

[0099] Based on the LightGBM-based account opening audit method shown in the above embodiment of the present invention, the embodiment of the present invention also discloses a LightGBM-based account opening audit device, such as Figure 4 As shown, it is a structural schematic diagram of an account opening review device based on LightGBM shown in an embodiment of the present invention, and the device includes:

[0100] The acquisition unit 401 is used to acquire the account opening material information collected by the bank account manager.

[0101] It should be noted that the account opening materials include at least the company legal person information, business address information and employee information collected on-site.

[0102] The data audit model 402 is used to use the account opening material information as the input of a pre-constructed data audit model, and predict the account opening material information based on the pre-constructed data audit model to obtain a prediction result, and the prediction result is used to indicate whether the account opening material information is qualified. The pre-constructed data audit model is constructed based on the construction unit 404. If the prediction result is qualified, the submission unit 403 is executed.

[0103] The submitting unit 403 is used to submit the account opening material information to the bank system so as to open a corresponding public account.

[0104] It should be noted that the specific principles and execution processes of each unit in the above-mentioned account opening review device based on LightGBM disclosed in the embodiments of the present invention are the same as those of the account opening review method based on LightGBM shown in the embodiments of the present invention. For the corresponding parts, reference can be made to the account opening review method based on LightGBM disclosed in the embodiments of the present invention, and details will not be elaborated here.

[0105] In the embodiments of the present invention, manual review is not required. The account opening material information is used as the input of a pre-constructed data review model, and a data review model constructed based on the LightGBM algorithm model is used to predict the account opening material information to obtain a prediction result, so as to determine whether the account opening material information is qualified. When it is determined that the prediction result is qualified, the account opening material information is submitted to the bank system to open a corresponding corporate account. Through the above method, the processing speed of opening a corporate account can be improved.

[0106] Based on the account opening review device based on LightGBM shown in the above embodiments of the present invention, in combination with Figure 4 , see Figure 5 , the account opening review device is further provided with an output unit 405.

[0107] The output unit 405 is configured to output a prompt message when it is determined that the prediction result is unqualified, so as to prompt the bank customer manager to re-collect the account opening material information.

[0108] It should be noted that after receiving the account opening material information re-collected by the bank customer manager, the execution returns to the acquisition unit 401.

[0109] In the embodiments of the present invention, manual review is not required. The account opening material information is used as the input of a pre-constructed data review model, and a data review model constructed based on the LightGBM algorithm model is used to predict the account opening material information to obtain a prediction result, so as to determine whether the account opening material information is qualified. When it is determined that the prediction result is qualified, the account opening material information is submitted to the bank system to open a corresponding corporate account. When it is determined that the prediction result is unqualified, a prompt message is output to prompt the bank customer manager to re-collect the account opening material information, and the execution returns to the acquisition unit. Through the above method, the processing speed of opening a corporate account can be improved.

[0110] Based on the account opening review device based on LightGBM shown in the above embodiments of the present invention, the construction unit 405 includes:

[0111] An acquisition module, configured to acquire sample data, where the sample data includes historical account opening material information that has passed the review within a historical time period.

[0112] An extraction module for processing the sample data to extract the sample feature data.

[0113] A training module for determining an initial LightGBM algorithm model and training the initial LightGBM algorithm model based on the sample feature data.

[0114] Optionally, the training module is specifically configured to: configure the initial network parameters of the initial LightGBM algorithm model based on the LightGBM algorithm; divide the sample feature data into a training set, a validation set, and a test set; train the initial network parameters using the training set and adjust the initial network parameters based on the validation set to obtain adjusted network parameters; and construct a LightGBM algorithm model based on the adjusted network parameters.

[0115] An evaluation module for evaluating the trained LightGBM algorithm model until the obtained evaluation result is within a preset range, and determining the currently trained LightGBM algorithm model as a data review model.

[0116] Optionally, the evaluation module is specifically configured to: evaluate the LightGBM algorithm model using the test set to obtain an evaluation result; if the evaluation result is within the preset range, determine the currently trained LightGBM algorithm model as a data review model; if the evaluation result is not within the predicted range, continue to train the currently trained LightGBM algorithm model based on the evaluation result, the training set, the validation set, and the test set.

[0117] In an embodiment of the present invention, sample data is obtained, the sample data is processed to extract sample feature data. An initial LightGBM algorithm model is determined, and the initial LightGBM algorithm model is trained based on the sample feature data. The trained LightGBM algorithm model is evaluated until the obtained evaluation result is within a preset range, and the currently trained LightGBM algorithm model is determined as a data review model. So as to subsequently use the constructed data review model to predict the account opening material information to obtain a prediction result to determine whether the account opening material information is qualified; when it is determined that the prediction result is qualified, the account opening material information is submitted to the banking system to open a corresponding corporate account. Through the above method, the processing speed of opening a corporate account can be improved.

[0118] An embodiment of the present invention also discloses an electronic device for running a database stored procedure, wherein when running the database stored procedure, the above-mentioned Figure 1 and Figure 2 disclosed account opening review method based on LightGBM is executed.

[0119] An embodiment of the present invention also discloses a computer storage medium, the storage medium including a stored database stored procedure, wherein when the database stored procedure runs, it controls the device where the storage medium is located to execute the above-mentioned Figure 1 and Figure 2 disclosed account opening audit method based on LightGBM.

[0120] In the context of the present disclosure, a computer storage medium may be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, 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), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0121] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments. In particular, for a system or system embodiment, since it is basically similar to a method embodiment, the description is relatively simple, and the relevant parts can refer to the description of the method embodiment. The systems and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative work.

[0122] Those skilled in the art can further realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0123] The foregoing description of the disclosed embodiments enables those skilled in the art to practice or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An account opening review method based on LightGBM, characterized in that, applied to opening a corporate account, the method includes: Obtain the account opening material information collected by the bank customer manager, and the account opening material information at least includes the company legal person information, business address information and employee situation information collected on-site; Use the account opening material information as the input of a pre-constructed data review model, and predict the account opening material information based on the pre-constructed data review model to obtain a prediction result, where the prediction result is used to indicate whether the account opening material information is qualified, and the pre-constructed data review model is constructed based on the LightGBM algorithm model; If the prediction result is qualified, submit the account opening material information to the bank system to open the corresponding corporate account; Among them, the pre-construction process of the data review model includes: Obtain sample data, and the sample data includes historical account opening material information that has passed the review within a historical time period; Process the sample data and extract the sample feature data; Determine the initial LightGBM algorithm model, and train the initial LightGBM algorithm model based on the sample feature data. Among them, input X and Y into the LightGBM model, and use the LightGBM algorithm to model the LightGBM model. X is the sample feature data, and Y is the review result. The review result 0 indicates that the review is qualified; the review result 1 indicates that the review is unqualified; Evaluate the trained LightGBM algorithm model until the obtained evaluation result is within the preset range, and determine the currently trained LightGBM algorithm model as the data review model.

2. The method according to claim 1, characterized in that, further includes: If the prediction result is unqualified, output a prompt message to prompt the bank customer manager to re-collect the account opening material information.

3. The method according to claim 1, characterized in that, The training of the initial LightGBM algorithm model based on the sample feature data includes: Based on the LightGBM algorithm, configure the initial network parameters of the initial LightGBM algorithm model; Divide the sample feature data into a training set, a validation set and a test set; Use the training set to train the initial network parameters, and adjust the initial network parameters based on the validation set to obtain the adjusted network parameters; Construct a LightGBM algorithm model based on the adjusted network parameters.

4. The method according to claim 1, characterized in that, The evaluation of the trained LightGBM algorithm model until the obtained evaluation result is within the preset range and determining the currently trained LightGBM algorithm model as the data review model includes: Use the test set to evaluate the LightGBM algorithm model to obtain an evaluation result; If the evaluation result is within the preset range, determine the current LightGBM algorithm model as the data review model; If the evaluation result is not within the predicted range, continue to train the current LightGBM algorithm model based on the evaluation result, the training set, the validation set, and the test set.

5. An account opening review device based on LightGBM, characterized in that the device includes: An acquisition unit for acquiring the account opening material information collected by the bank customer manager, where the account opening material information at least includes the company legal person information, business address information, and employee situation information collected on-site; A data review model for using the account opening material information as the input of a pre-constructed data review model, and predicting the account opening material information based on the pre-constructed data review model to obtain a prediction result, where the prediction result is used to indicate whether the account opening material information is qualified. The pre-constructed data review model is constructed by a construction unit. If the prediction result is qualified, the submission unit is executed; The submission unit for submitting the account opening material information to the bank system for opening a corresponding corporate account; Among them, the construction unit includes: An acquisition module for acquiring sample data, where the sample data includes historical account opening material information that has passed the review within a historical time period; An extraction module for processing the sample data and extracting the sample feature data; A training module for determining an initial LightGBM algorithm model and training the initial LightGBM algorithm model based on the sample feature data; among them, input X and Y into the LightGBM model, and use the LightGBM algorithm to model the LightGBM model. X is the sample feature data, and Y is the review result. The review result 0 indicates that the review is qualified; the review result 1 indicates that the review is unqualified; An evaluation module for evaluating the trained LightGBM algorithm model until the obtained evaluation result is within a preset range, and determining the currently trained LightGBM algorithm model as the data review model.

6. The device according to claim 5, characterized in that it further includes: An output unit for outputting a prompt message to prompt the bank customer manager to re-collect the account opening material information when it is determined that the prediction result is unqualified.

7. An electronic device, characterized in that the electronic device is used to run a program, where when the program runs, it executes the LightGBM-based account opening review method described in any one of claims 1-4.

8. A computer storage medium, characterized in that the storage medium includes a stored program, where when the program runs, it controls the device where the storage medium is located to execute the LightGBM-based account opening review method described in any one of claims 1-4.

Images