Customer classification method and device, computer device, storage medium and program product

Abstract

The application relates to a customer classification method and device, computer equipment, a storage medium and a program product, and relates to the technical field of artificial intelligence. The method comprises the following steps: obtaining target attribute information of a target customer; determining an ordered reinforced average distance between the target customer and each clustering center according to the target attribute information and the clustering center corresponding to each preset customer type; and determining a customer type of the target customer according to the ordered reinforced average distance between the target customer and each clustering center. The method can recommend a bank product meeting the demand of a user, and ensure that the recommended bank product matches the customer type.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a customer classification method, apparatus, computer equipment, storage medium, and program product. Background Technology

[0002] With the booming development of the financial services industry, a wide variety of banking products have emerged for customers to choose from. Product recommendation involves recommending suitable products to customers based on existing banking product resources and their needs or expectations.

[0003] Traditionally, the presentation of bank products relies on display boards and manual push notifications to attract customers' attention, combined with on-site inquiries from bank staff to recommend bank products.

[0004] However, the recommended products may fail to meet the user's needs. Therefore, how to recommend products that meet the user's needs has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] Based on this, it is necessary to provide a customer classification method, device, computer equipment, storage medium, and program product that can recommend bank products that meet user needs and ensure that the recommended bank products match the customer type, in order to address the above-mentioned technical problems.

[0006] Firstly, this application provides a customer classification method. The method includes:

[0007] Obtain target attribute information of target customers;

[0008] Based on the target attribute information and the cluster centers corresponding to each preset customer type, the ordered reinforced average distance between the target customer and each of the cluster centers is determined;

[0009] The customer type of the target customer is determined based on the ordered reinforced average distance between the target customer and each of the cluster centers.

[0010] In one embodiment, determining the ordered reinforced average distance between the target customer and each of the cluster centers based on the target attribute information and the cluster centers corresponding to each preset customer type includes:

[0011] For each cluster center, the distance between each sub-attribute information and the corresponding sub-cluster center is determined based on each sub-attribute information in the target attribute information and the corresponding sub-cluster center;

[0012] Based on the distance between each of the sub-attribute information and the corresponding sub-cluster center, the ordered reinforced average distance between the target customer and the cluster center is determined.

[0013] In one embodiment, determining the distance between each sub-attribute information and its corresponding sub-cluster center based on each sub-attribute information and its corresponding sub-cluster center in the target attribute information includes:

[0014] Determine the absolute value of the first difference between each of the sub-attribute information and the corresponding sub-cluster center;

[0015] For each of the sub-attribute information, determine the maximum value from the sub-attribute information and the corresponding sub-cluster center;

[0016] The distance between the sub-attribute information and the corresponding sub-cluster center is determined based on the ratio of the absolute value to the maximum value.

[0017] In one embodiment, determining the distance between the sub-attribute information and the corresponding sub-cluster center based on the ratio of the absolute value to the maximum value includes:

[0018] The ratio is used as the distance between the sub-attribute information and the corresponding sub-cluster center.

[0019] In one embodiment, determining the ordered reinforced average distance between the target customer and the cluster center based on the distance between each of the sub-attribute information and the corresponding sub-cluster center includes:

[0020] Determine the product between each distance and its corresponding target weight; the target weight is determined according to the target ordered augmentation average model, which is obtained by training a preset initial ordered augmentation average model based on the training set corresponding to the preset customer type;

[0021] The summation result is obtained by summing the products of each distance and its corresponding target weight.

[0022] The summation result is used as the ordered reinforced average distance between the target customer and the cluster center.

[0023] In one embodiment, determining the customer type of the target customer based on the ordered reinforced average distance between the target customer and each of the cluster centers includes:

[0024] The preset customer type corresponding to the target cluster center is taken as the customer type of the target customer; the target cluster center is the cluster center corresponding to the smallest ordered reinforced average distance.

[0025] In one embodiment, the method further includes:

[0026] Obtain the training set corresponding to the preset customer type;

[0027] Based on the training set corresponding to the preset customer type and the preset customer type, the initial ordered augmented average model is trained to obtain the target ordered augmented average model.

[0028] In one embodiment, the method further includes:

[0029] Obtain attribute information samples of each customer in the training set corresponding to each of the preset customer types;

[0030] For each of the preset customer types, the sub-cluster centers corresponding to each preset customer type are determined based on the attribute information samples of the same type of each customer in the training set corresponding to the preset customer type.

[0031] Based on the sub-cluster centers corresponding to the preset customer type, determine the cluster center corresponding to the preset customer type.

[0032] Secondly, this application also provides a customer classification device. The device includes:

[0033] The first acquisition module is used to acquire target attribute information of target customers;

[0034] The first determining module is used to determine the ordered reinforced average distance between the target customer and each of the cluster centers based on the target attribute information and the cluster centers corresponding to each preset customer type.

[0035] The second determining module is used to determine the customer type of the target customer based on the ordered reinforced average distance between the target customer and each of the cluster centers.

[0036] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps:

[0037] Obtain target attribute information of target customers;

[0038] Based on the target attribute information and the cluster centers corresponding to each preset customer type, the ordered reinforced average distance between the target customer and each of the cluster centers is determined;

[0039] The customer type of the target customer is determined based on the ordered reinforced average distance between the target customer and each of the cluster centers.

[0040] Fourthly, this application also provides a computer-readable storage medium. The aforementioned computer-readable storage medium stores a computer program thereon, which, when executed by a processor, performs the following steps:

[0041] Obtain target attribute information of target customers;

[0042] Based on the target attribute information and the cluster centers corresponding to each preset customer type, the ordered reinforced average distance between the target customer and each of the cluster centers is determined;

[0043] The customer type of the target customer is determined based on the ordered reinforced average distance between the target customer and each of the cluster centers.

[0044] Fifthly, this application also provides a computer program product. The aforementioned computer program product includes a computer program that, when executed by a processor, performs the following steps:

[0045] Obtain target attribute information of target customers;

[0046] Based on the target attribute information and the cluster centers corresponding to each preset customer type, the ordered reinforced average distance between the target customer and each of the cluster centers is determined;

[0047] The customer type of the target customer is determined based on the ordered reinforced average distance between the target customer and each of the cluster centers.

[0048] The aforementioned customer classification method, apparatus, computer equipment, storage medium, and program product acquire target attribute information of target customers, determine the ordered reinforced average distance between the target customer and each of the cluster centers corresponding to each preset customer type based on the target attribute information and the cluster centers, and determine the customer type of the target customer based on the ordered reinforced average distance between the target customer and each of the cluster centers. Traditional methods use display boards and manual recommendations to push marketing products to customers; however, due to the ambiguity in understanding user needs during the push process, the pushed products cannot meet user needs. This application's embodiment introduces an ordered reinforced average model, considering the target attribute information of target customers, and accurately classifies customers based on the ordered reinforced average distance between the target customer's target attribute information and the cluster centers of the preset customer types. This enables the recommendation of bank products that meet user needs based on customer type, ensuring that the recommended bank products match the customer type. Attached Figure Description

[0049] Figure 1 This is a diagram illustrating the application environment of a customer classification method in one embodiment.

[0050] Figure 2 This is a flowchart illustrating a customer classification method in one embodiment;

[0051] Figure 3 This is one of the flowcharts illustrating a method for determining the ordered reinforcement average distance in one embodiment;

[0052] Figure 4 This is a flowchart illustrating a method for determining the distance between each sub-attribute information and its corresponding sub-cluster center in one embodiment.

[0053] Figure 5 This is a second flowchart illustrating the method for determining the ordered reinforcement average distance in one embodiment;

[0054] Figure 6 This is a flowchart illustrating the training method of an ordered augmentation averaging model in one embodiment.

[0055] Figure 7 This is a flowchart illustrating a method for determining cluster centers in one embodiment;

[0056] Figure 8 This is a structural block diagram of a customer sorting device in one embodiment;

[0057] Figure 9 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0059] This application provides a customer classification method that can be applied to, for example... Figure 1 The application environment shown. Figure 1 This is a diagram illustrating the application environment of a customer classification method in one embodiment. The application environment includes a terminal 102 and a server 104. The terminal 102 communicates with the server 104 via a network to obtain target attribute information of target customers; based on the target attribute information and the cluster centers corresponding to each preset customer type, it determines the ordered reinforced average distance between the target customer and each cluster center; based on the ordered reinforced average distance between the target customer and each cluster center, it determines the customer type of the target customer. A data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104 or placed on the cloud or other network servers. The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can be smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. Portable wearable devices can be smartwatches, smart bracelets, head-mounted devices, etc. The computer device can also be a server.

[0060] In one embodiment, such as Figure 2 As shown, Figure 2 This application provides a flowchart illustrating a customer classification method, which is applied to... Figure 1 Taking terminal 102 as an example, the following steps are included:

[0061] S201, Obtain target attribute information of target customers.

[0062] The target attribute information refers to the data attribute information after standardizing the original attribute information of the target customers. The original attribute information may include textual attribute information of the target customers, such as gender, education level, occupation, etc.; or it may include data attribute information of the target customers, such as age, frequency of purchasing bank products, etc.

[0063] Specifically, when the original attribute information of the target customer includes text attribute information, the text attribute information is converted into data attribute information, and then standardized to obtain standardized data attribute information. Then, the target customer's target attribute information is obtained based on the standardized data attribute information. When the original attribute information of the target customer includes data attribute information, the original attribute information is directly standardized to obtain standardized attribute information, and the target customer's target attribute information is obtained based on the standardized attribute information. Standardization refers to scaling the data proportionally to ensure that each set of data falls within a specific interval while maintaining the data distribution. In this embodiment, all target attributes are standardized so that the order of magnitude of each target attribute falls within the same interval, for example, all within [0,1].

[0064] It should be noted that the magnitudes of the target attribute information differ. For example, the target attribute information corresponding to age is 40, and the target attribute information corresponding to the frequency of purchasing banking products is 4. Because the ordered reinforced average distance between these target attribute information and the cluster centers is more significantly influenced by the age attribute in subsequent step S202, the calculation accuracy of the ordered reinforced average distance between the target attribute information and the cluster centers is low. For instance, if a target customer includes two attributes, age and the frequency of purchasing banking products, with each attribute having a weight of 0.5, then the ordered reinforced average distance is equal to 40*0.5 + 4*0.5, which is 22. Therefore, the ordered reinforced average distance of 22 is closer to the target attribute information corresponding to age (40). Thus, when different attribute information has the same weight, the ordered reinforced average distance is significantly closer to the attribute information corresponding to the larger magnitude. Therefore, standardizing the target attribute information of target customers can balance the impact of each target attribute information on subsequent customer type judgments.

[0065] Specifically, in this embodiment, the acquisition method is to capture the facial information of the target customer through a camera, obtain the attribute information of all customers in the database, and obtain the target attribute information of the target customer from the attribute information of all customers based on the facial recognition results.

[0066] S202, Based on the target attribute information and the cluster centers corresponding to each preset customer type, determine the ordered reinforced average distance between the target customer and each cluster center.

[0067] In this system, cluster centers represent attribute information of a preset customer type, and each cluster center contains multiple sub-cluster centers. Correspondingly, target attribute information represents attribute information of the target customer, and target attribute information contains multiple sub-attributes. The number of sub-cluster centers is the same as the number of sub-attributes, and the distance between a sub-cluster center and a sub-attribute is called the sub-distance. The ordered reinforced average distance represents the difference between the target customer and the cluster center; the smaller the ordered reinforced average distance, the closer the target customer's type is to the preset customer type corresponding to the cluster center.

[0068] Optionally, firstly, weights can be set for each sub-cluster center. The absolute value of the difference between each sub-attribute information and each sub-cluster center is multiplied by the weight corresponding to each sub-cluster center, and the resulting product is used as the new absolute value for each sub-cluster center. Next, based on the ratio of the new absolute value for each sub-cluster center to the maximum value of each sub-attribute information among all sub-cluster centers, the sub-distance between each sub-attribute information and its corresponding sub-cluster center is obtained. Finally, based on the sum of all sub-distances, the ordered reinforced average distance between the target customer and each cluster center is obtained.

[0069] Optionally, firstly, weights can be set for each sub-cluster center. Next, the ratio of the absolute value of the difference between each sub-attribute information and each sub-cluster center to the maximum value of each sub-attribute information within each sub-cluster center is used as the sub-distance. Then, the product of the weights of each sub-cluster center and each sub-distance is used as the new sub-distance. Finally, the sum of these new sub-distances is used as the ordered reinforced average distance between the target customer and each cluster center.

[0070] In this embodiment, the ordered enhanced average distance is used as the basis for determining the customer type. This not only retains the distance between each sub-attribute information of the target customer and each sub-cluster center, but also combines the influence of different sub-attribute information on each sub-cluster center, which can effectively characterize the distance between the target customer and the cluster center.

[0071] S203. Determine the customer type of the target customer based on the ordered reinforced average distance between the target customer and each cluster center.

[0072] Specifically, after obtaining the ordered reinforced average distance between the target customer and each cluster center, the minimum ordered reinforced average distance is determined, and the preset customer type of the cluster center corresponding to the minimum ordered reinforced average distance is taken as the customer type of the target customer.

[0073] Optionally, if the target customer has the same ordered reinforced average distance as multiple different cluster centers, different corresponding coefficients can be set for different cluster centers. Based on the corresponding coefficients of the cluster centers and the ordered reinforced average distance between the target customer and the cluster centers, a new ordered reinforced average distance can be obtained. Then, based on the new ordered reinforced average distance, the customer type of the target customer can be determined.

[0074] For example, if there are two cluster centers, and the ordered reinforced average distance between the target customer and cluster center 1 is 20, with a corresponding coefficient of 0.6, then the new ordered reinforced average distance between the target customer and cluster center 1 is 12. Similarly, if the ordered reinforced average distance between the target customer and cluster center 2 is 20, with a corresponding coefficient of 0.4, then the new ordered reinforced average distance between the target customer and cluster center 2 is 8. Clearly, the ordered reinforced average distance between the target customer and cluster center 2 is smaller, thus determining the target customer type as the preset customer type corresponding to cluster center 2.

[0075] The customer classification method provided in this application obtains the target customer's target attribute information, determines the ordered reinforced average distance between the target customer and each cluster center based on the target attribute information and the cluster centers corresponding to each preset customer type, and determines the target customer's customer type based on the ordered reinforced average distance between the target customer and each cluster center. Traditional methods use display boards and manual recommendations to push marketing products to customers. However, due to the vague understanding of user needs during the push process, the pushed products cannot meet the user's needs. This application introduces an ordered reinforced average model, considering the target customer's target attribute information, and accurately classifies customers based on the ordered reinforced average distance between the target customer's target attribute information and the cluster centers of the preset customer types. This enables the recommendation of bank products that meet the user's needs based on the customer type, ensuring that the recommended bank products match the customer type.

[0076] In one optional embodiment of this application, such as Figure 3 As shown, Figure 3 This is a flowchart illustrating one embodiment of a method for determining ordered reinforced average distance. This embodiment relates to an optional implementation of "how to determine the ordered reinforced average distance between a target customer and each cluster center based on target attribute information and the cluster centers corresponding to each preset customer type." Based on the above embodiment, S202 includes:

[0077] S301, For each cluster center, determine the distance between each sub-attribute information and the corresponding sub-cluster center based on each sub-attribute information in the target attribute information and the corresponding sub-cluster center.

[0078] In this target attribute information, each sub-attribute is categorized according to its type. The number of sub-attributes in the target attribute information is the same as the number of sub-cluster centers in each cluster center, and the type of each sub-attribute in the target attribute is consistent with the type of its corresponding sub-cluster center.

[0079] Specifically, firstly, based on the sub-attribute information in the target attribute information, the sub-cluster centers corresponding to each cluster center are obtained. Then, based on the sub-attribute information and the corresponding sub-cluster centers in the target attribute information, the distance between each sub-attribute information and its corresponding sub-cluster center is obtained.

[0080] S302, Based on the distance between each sub-attribute information and the corresponding sub-cluster center, determine the ordered reinforced average distance between the target customer and the cluster center.

[0081] Specifically, the distance between each obtained sub-attribute information and its corresponding sub-cluster center is used as a sub-distance. A weight is assigned to each sub-distance, and the number of weights is the same as the number of sub-distances. The sum of the products of the weights and each sub-distance is used as the ordered reinforced average distance. For example, if each sub-attribute information has the same influence on the judgment of the target customer type, then the weight of each sub-distance is set to 1, and the ordered reinforced average distance between the target customer and the cluster center is the sum of all sub-distances.

[0082] In this embodiment, the customer attribute information is finely divided into a set. While considering the comprehensiveness of the customer attribute information, the impact of excessively large or small data in a single customer attribute on the acquisition of ordered reinforced average distance is balanced, thereby improving the reliability of ordered reinforced average distance between the target customer and each cluster center.

[0083] In one optional embodiment of this application, such as Figure 4 As shown, Figure 4 This is a flowchart illustrating a method for determining the distance between each sub-attribute information and its corresponding sub-cluster center in one embodiment. This embodiment relates to an optional implementation of "how to determine the distance between each sub-attribute information and its corresponding sub-cluster center based on each sub-attribute information in the target attribute information and its corresponding sub-cluster center". Based on the above embodiment, S301 includes:

[0084] S401, determine the absolute value of the first difference between each sub-attribute information and the corresponding sub-cluster center.

[0085] It should be noted that using the absolute value of the first difference between each sub-attribute information and the corresponding sub-cluster center as the data result can still accurately obtain the distance between the sub-attribute information and the corresponding sub-cluster center based on S403 below, even when the distance between the sub-attribute information and the sub-cluster center is negative.

[0086] For example, the sub-attribute information of the target customer is x i The corresponding sub-cluster center is y i Then the absolute value of the first difference between the sub-attribute information and the sub-cluster center is |x i -y i |or|y i -x i |

[0087] S402, for each sub-attribute information, determine the maximum value from the sub-attribute information and the corresponding sub-cluster centers.

[0088] It should be noted that both sub-attribute information and sub-cluster centers are expressions of information attributes, therefore all values ​​are non-negative, meaning the maximum value is also non-negative.

[0089] For example, if the sub-attribute information x i If the value is greater than the corresponding sub-cluster center yi, then retrieve the sub-attribute information x. i This serves as the maximum value for subsequent sub-distance calculations; if the sub-attribute information x i If the distance is less than the corresponding sub-cluster center yi, then the sub-cluster center yi is taken as the maximum value for subsequent sub-distance calculations.

[0090] S403, determine the distance between the sub-attribute information and the corresponding sub-cluster center based on the ratio of the absolute value to the maximum value.

[0091] It should be noted that since the values ​​of S402 are all non-negative, and are used as the denominator in this step, when the sub-attribute information x... i and the corresponding sub-cluster centers y i When both are 0, based on the sub-attribute information x i and the corresponding sub-cluster centers y i If the absolute value of the first difference is 0, then the distance between the sub-attribute information and the corresponding sub-cluster center is 0. Therefore, this step directly determines that the distance between the sub-attribute information and the corresponding sub-cluster center is 0. i and the corresponding sub-cluster centers y i This is performed when none of the values ​​is zero.

[0092] In this embodiment of the application, the distance between each sub-attribute information and each cluster center is obtained based on the corresponding sub-cluster center or the sub-attribute information itself, which can effectively obtain the distance between the sub-attribute information and the corresponding sub-cluster center.

[0093] In an optional embodiment of this application, the above-described S403, which determines the distance between the sub-attribute information and the corresponding sub-cluster center based on the ratio of the absolute value to the maximum value, further includes:

[0094] The ratio is used as the distance between the sub-attribute information and the corresponding sub-cluster center.

[0095] For example, after obtaining the absolute value and maximum value, if neither the sub-attribute information nor the corresponding sub-cluster center is 0, then the distance between the sub-attribute information and the corresponding sub-cluster center is represented as:

[0096]

[0097] Where, d r (x i, y i ) represents the distance between the i-th sub-attribute information and the corresponding sub-cluster center, x i For the target customer's sub-attribute information, y i For the corresponding sub-cluster centers, max{x i ,y i} represents taking x i y i The maximum value.

[0098] In this embodiment, the ratio is used as the distance between the sub-attribute information and the corresponding sub-cluster center. This preserves the original features of the sub-attribute information or sub-cluster center while also obtaining the relationship between the sub-attribute information and the sub-cluster center. Therefore, the distance between the sub-attribute information and the corresponding sub-cluster center can be obtained more accurately.

[0099] In one optional embodiment of this application, such as Figure 5 As shown, Figure 5 This is a second flowchart illustrating a method for determining the ordered reinforced average distance in one embodiment. This embodiment relates to an optional implementation of "how to determine the ordered reinforced average distance between the target customer and the cluster center based on the distance between each sub-attribute information and the corresponding sub-cluster center". Based on the above embodiment, S302 includes:

[0100] S501, determine the product result between each distance and the corresponding target weight; the target weight is determined according to the target ordered reinforcement average model, which is obtained by training the preset initial ordered reinforcement average model based on the training set corresponding to the preset customer type.

[0101] Among them, the ordered reinforced average model is a model for determining customer type based on the ordered reinforced average distance. The ordered reinforced average model is used to classify target customers and determine the distance between the target customer and a certain cluster center. The closer the distance, the closer the target customer's type is to the preset customer type corresponding to that cluster center.

[0102] Specifically, the parameters of the initial ordered reinforced average model are set to random initialization and the parameter update method is gradient descent. The ordered reinforced average model is trained based on the customer attribute information and customer type in the training set. The trained model parameters include target weights corresponding to the customer attribute information.

[0103] For example, if the customer attribute information contains n attributes, the target weight is represented as follows:

[0104] OWA(x1,...,x n )={w1,...,w n (Equation 2)

[0105] Where, x i w represents the i-th attribute information. i The weight corresponding to the i-th attribute information is in the range [0,1], and the sum of all weight factors is 1.

[0106] S502, sum the products of each distance and the corresponding target weight to obtain the summation result.

[0107] Specifically, the updated distances are obtained by multiplying each distance by its corresponding target weight, and then the updated distances are summed. For example, the sum of the products of the target customer and a certain cluster center is:

[0108]

[0109] Where X = {x1,...,x} n Let} represent n information attributes of the target customer, Y = (y1,...,y2) n ) represents n information attributes of a preset customer type.

[0110] S503 uses the summation result as the ordered reinforced average distance between the target customer and the cluster center.

[0111] In the embodiments of this application, by assigning low (or high) weights, the influence of excessively large or small deviations in the distance between sub-attribute information and the corresponding sub-cluster centers is reduced (or strengthened) to obtain a more reliable ordered strengthened average distance and provide a more accurate basis for customer classification.

[0112] In an optional embodiment of this application, S203 is further described, which involves determining the customer type of the target customer based on the ordered reinforced average distance between the target customer and each cluster center, including:

[0113] The preset customer type corresponding to the target cluster center is taken as the customer type of the target customer; the target cluster center is the cluster center corresponding to the smallest ordered reinforced average distance.

[0114] It should be noted that the target cluster center refers to one of the cluster centers corresponding to each preset customer type. For a target customer, the number of ordered reinforced average distances is the same as the number of preset customer types. For example, if there are m preset customer types, to determine the type of a target customer, it is necessary to calculate the m ordered reinforced average distances between the target customer and the cluster centers of the m preset customer types. The preset customer type corresponding to the cluster center with the smallest ordered reinforced average distance is the target customer's customer type.

[0115] Specifically, when a target customer has the same ordered reinforced average distance to two different cluster centers, the preset customer type corresponding to one of the cluster centers is selected as the target customer's customer type.

[0116] In the embodiments of this application, the distance between the target customer and the cluster centers of different preset customer types is compared by ordered reinforced average distance. Based on the minimum ordered reinforced average distance, the customer type is determined, thereby improving the accuracy of customer classification.

[0117] In one optional embodiment of this application, such as Figure 6 As shown, Figure 6 A flowchart illustrating a training method for an ordered augmented averaging model provided in this application embodiment includes the following steps:

[0118] S601, Obtain the training set corresponding to the preset customer type.

[0119] The training set for each preset customer type includes the attribute information of all customers of that preset customer type; that is, each preset customer type corresponds to one training set.

[0120] S602, based on the training set corresponding to the preset customer type and the preset customer type, train the initial ordered augmented average model to obtain the target ordered augmented average model.

[0121] The ordered reinforcement averaging model is trained on the training set. Training stops when the model error between the predicted and actual results meets an error threshold. If the model error is less than the threshold, the training meets the accuracy requirements, and training stops.

[0122] Specifically, the training method is as follows: First, for the training set corresponding to the preset customer type, an initial ordered augmented average model is trained based on the attribute information and customer type of all customers in the training set, obtaining an ordered augmented average model containing training weights. Next, the attribute information of any customer in the training set is fed into the trained ordered augmented average model, and the predicted customer type is output. The model error between the predicted customer type and the actual customer type is obtained. Finally, an error threshold is set. When the model error is less than the error threshold, training stops, and a target ordered augmented average model containing the target weights is obtained.

[0123] Optionally, the customer information attributes in the training set can be divided in a 3:1 ratio. Based on 3 / 4 of the customer information attributes and their corresponding customer types, an initial ordered augmented average model can be trained to obtain an ordered augmented average model containing training weights. Next, the customer information attributes of the remaining 1 / 4 are fed into the trained ordered augmented average model, outputting the predicted customer types for that 1 / 4 of the customers. Then, based on the actual customer types of the 1 / 4 customers and the predicted customer types of the 1 / 4 customers, the mean model error is obtained. Finally, the model error is compared with an error threshold.

[0124] In the embodiments of this application, customer attribute information and corresponding customer types are used as the training set, resulting in more accurate model parameters and faster training speed. Furthermore, the error threshold can be adjusted autonomously according to actual conditions, offering high flexibility.

[0125] In one optional embodiment of this application, such as Figure 7 As shown, Figure 7 A flowchart illustrating a method for determining cluster centers provided in this application embodiment includes the following steps:

[0126] S701, Obtain attribute information samples of each customer in the training set corresponding to each preset customer type.

[0127] The customer attribute information sample includes attribute information for multiple customer types and corresponding customer type information. The order of attribute information for each customer is sorted according to a uniform type, and the training set corresponding to each preset customer type is a subset of the entire training set. For example, if there are p preset customer types, there are p corresponding training subsets, and the customer type information corresponding to each training subset is consistent.

[0128] S702, for each preset customer type, determine the sub-cluster centers corresponding to each preset customer type based on the attribute information samples of the same type of each customer in the training set corresponding to the preset customer type.

[0129] Specifically, based on the training set corresponding to each preset customer type, the attribute information of all customers in the training set is listed in a unified type order, and the mean of the attribute information samples of the same type for each customer is calculated as the sub-cluster center. In other words, the number of sub-cluster centers is the same as the number of attribute information types in the customer attribute information samples.

[0130] For example, if the training set corresponding to the k-th preset customer type has m customers, and each customer has n attribute information, then the corresponding n-th sub-cluster center is represented as:

[0131]

[0132] Where k represents the category of the preset customer type.

[0133] S703, determine the cluster center corresponding to the preset customer type based on the sub-cluster centers corresponding to the preset customer type.

[0134] In this system, the cluster centers corresponding to the preset customer types are sets of all sub-cluster centers, and the attribute information of the cluster centers is consistent with the order of the customer's attribute information. For example, if there are k preset customer types and each customer has n attribute information, then there are k cluster centers, and each cluster center contains n sub-cluster centers.

[0135] In the embodiments of this application, sub-cluster centers are calculated using the average information attributes of all customers of the same customer type. The accuracy of the sub-cluster centers is positively correlated with the number of customers of that customer type; that is, the more numerous the customers corresponding to that customer type, the more accurate the obtained sub-cluster centers. In other words, given a bank branch with a sufficient number of customers, the more accurate the values ​​of the obtained sub-cluster centers, the more reliable the resulting cluster centers.

[0136] In one specific embodiment, a customer classification method is provided, including:

[0137] The first step is to capture the customer's arrival information when the customer arrives at the store using a binocular camera. Based on the facial recognition results of the target customer, the camera retrieves the facial information and corresponding attribute information of all customers.

[0138] The second step is to obtain n attribute information of the target customer based on the facial recognition matching results, such as gender, age, height, whether or not they have purchased marketing products. Among these, the information attributes include text information attributes and first data information attributes.

[0139] The third step is to convert the target customer's text information attributes into second data information attributes. For example, in the customer's gender attribute, male is converted into data 1 and female is converted into data 2; in the customer's whether they have purchased marketing products attribute, male is converted into data 1 and female is converted into data 2, etc.

[0140] The fourth step is to standardize the first and second data information attributes.

[0141] The fifth step is to divide the customer dataset consisting of all customers into a training set and a test set, train an ordered reinforced average model, and obtain the target weights.

[0142] The sixth step is to divide the training set into p subsets, each subset corresponding to a preset customer type, and calculate the cluster center for each preset customer type.

[0143] Step 7: For each cluster center, calculate the distance between the target customer's sub-attribute information and the sub-cluster centers of the cluster center;

[0144] Step 8: Calculate the ordered reinforcement average distance based on the target weight, the sub-attribute information of the target customer, and the distance between the sub-cluster centers of the cluster center;

[0145] Step 9: Based on the ordered enhanced average distance between the target customer and multiple cluster centers, obtain the ordered enhanced average distance between the target customer and different preset customer types;

[0146] Step 10: For multiple ordered reinforced average distances, determine the smallest ordered reinforced average distance, and determine the preset customer type of the cluster center corresponding to the smallest ordered reinforced average distance as the target customer type.

[0147] The aforementioned customer classification method obtains the target customer's target attribute information, determines the ordered reinforced average distance between the target customer and each cluster center based on the target attribute information and the cluster centers corresponding to each preset customer type, and then determines the target customer's customer type based on the ordered reinforced average distance between the target customer and each cluster center. Traditional methods use display boards and manual recommendations to push marketing products to customers. However, due to the vague understanding of user needs during the push process, the pushed products cannot meet the user's needs. This customer classification method introduces an ordered reinforced average model, considers the target customer's target attribute information, and accurately classifies customers based on the ordered reinforced average distance between the target customer's target attribute information and the cluster centers of the preset customer type. This allows for the recommendation of bank products that meet the user's needs based on their customer type, ensuring that the recommended bank products match the customer type.

[0148] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0149] Based on the same inventive concept, this application also provides a customer classification device for implementing the customer classification method described above. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations in one or more customer classification device embodiments provided below can be found in the limitations of the customer classification method described above, and will not be repeated here.

[0150] In one embodiment, such as Figure 8 As shown, a customer classification device 800 is provided. The device includes: a first acquisition module 801, a first determination module 802, and a second determination module 803, wherein:

[0151] The first acquisition module 801 is used to acquire target attribute information of the target customer;

[0152] The first determining module 802 is used to determine the ordered reinforced average distance between the target customer and each cluster center based on the target attribute information and the cluster centers corresponding to each preset customer type.

[0153] The second determining module 803 is used to determine the customer type of the target customer based on the ordered reinforced average distance between the target customer and each cluster center.

[0154] In one embodiment, the first determining module 802 further includes:

[0155] The first determining unit is used to determine the distance between each sub-attribute information and the corresponding sub-cluster center for each cluster center, based on each sub-attribute information in the target attribute information and the corresponding sub-cluster center.

[0156] The second determining unit is used to determine the ordered reinforced average distance between the target customer and the cluster center based on the distance between each sub-attribute information and the corresponding sub-cluster center.

[0157] In one embodiment, the first determining unit is specifically configured to determine the absolute value of a first difference between each sub-attribute information and its corresponding sub-cluster center; for each sub-attribute information, determine the maximum value from the sub-attribute information and its corresponding sub-cluster center; determine the distance between the sub-attribute information and its corresponding sub-cluster center based on the ratio of the absolute value to the maximum value; and use the ratio as the distance between the sub-attribute information and its corresponding sub-cluster center.

[0158] In one embodiment, the second determining unit is specifically used to determine the product result between each distance and the corresponding target weight; the target weight is determined according to the target ordered strengthening average model, which is obtained by training a preset initial ordered strengthening average model based on the training set corresponding to the preset customer type; the product results between each distance and the corresponding target weight are summed to obtain a summation result; the summation result is used as the ordered strengthening average distance between the target customer and the cluster center.

[0159] In one embodiment, the second determining module 803 further includes:

[0160] The judgment unit uses the preset customer type corresponding to the target cluster center as the customer type of the target customer; the target cluster center is the cluster center corresponding to the smallest ordered reinforced average distance.

[0161] In one embodiment, the customer sorting device 800 may further include:

[0162] The second acquisition module is used to acquire the training set corresponding to the preset customer type;

[0163] The training module is used to train the initial ordered augmented average model to obtain the target ordered augmented average model based on the training set corresponding to the preset customer type and the preset customer type.

[0164] In one embodiment, the customer sorting device 800 may further include:

[0165] The third acquisition module is used to acquire attribute information samples of each customer in the training set corresponding to each preset customer type.

[0166] The third determination module is used to determine the sub-cluster centers corresponding to each preset customer type based on the attribute information samples of the same type of each customer in the training set corresponding to the preset customer type.

[0167] The fourth determination module is used to determine the cluster center corresponding to the preset customer type based on the sub-cluster centers corresponding to the preset customer type.

[0168] Each module in the aforementioned customer classification device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0169] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 9 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores customer attribute information data. The network interface communicates with external terminals via a network connection. When executed by the processor, the computer program implements a customer classification method.

[0170] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 9 As shown, the computer device includes a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When executed by the processor, the computer program implements a customer classification method. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device casing, or an external keyboard, touchpad, or mouse.

[0171] Those skilled in the art will understand that Figure 9 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0172] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to perform the following steps:

[0173] Obtain target attribute information of target customers;

[0174] Based on the target attribute information and the cluster centers corresponding to each preset customer type, determine the ordered reinforced average distance between the target customer and each cluster center;

[0175] The customer type of the target customer is determined based on the ordered reinforced average distance between the target customer and each cluster center.

[0176] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0177] For each cluster center, the distance between each sub-attribute information and its corresponding sub-cluster center is determined based on each sub-attribute information in the target attribute information and the corresponding sub-cluster center.

[0178] Based on the distance between each sub-attribute information and the corresponding sub-cluster center, the ordered reinforced average distance between the target customer and the cluster center is determined.

[0179] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0180] Determine the absolute value of the first difference between each sub-attribute information and its corresponding sub-cluster center;

[0181] For each sub-attribute information, determine the maximum value from the sub-attribute information and the corresponding sub-cluster centers;

[0182] The distance between the sub-attribute information and the corresponding sub-cluster center is determined by the ratio of the absolute value to the maximum value.

[0183] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0184] The ratio is used as the distance between the sub-attribute information and the corresponding sub-cluster center.

[0185] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0186] Determine the product between each distance and the corresponding target weight; the target weight is determined based on the target ordered augmentation average model, which is obtained by training a preset initial ordered augmentation average model on a training set corresponding to a preset customer type;

[0187] The summation result is obtained by summing the products of each distance and its corresponding target weight.

[0188] The summation result is used as the ordered reinforced average distance between the target customer and the cluster center.

[0189] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0190] The preset customer type corresponding to the target cluster center is taken as the customer type of the target customer; the target cluster center is the cluster center corresponding to the smallest ordered reinforced average distance.

[0191] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0192] Obtain the training set corresponding to the preset customer type;

[0193] Based on the training set corresponding to the preset customer type and the preset customer type, the initial ordered augmented average model is trained to obtain the target ordered augmented average model.

[0194] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0195] Obtain attribute information samples of each customer in the training set corresponding to each preset customer type.

[0196] For each preset customer type, the sub-cluster centers corresponding to each preset customer type are determined based on the attribute information samples of the same type of customers in the training set corresponding to the preset customer type.

[0197] Based on the sub-cluster centers corresponding to the preset customer types, determine the cluster centers corresponding to the preset customer types.

[0198] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0199] Obtain target attribute information of target customers;

[0200] Based on the target attribute information and the cluster centers corresponding to each preset customer type, determine the ordered reinforced average distance between the target customer and each cluster center;

[0201] The customer type of the target customer is determined based on the ordered reinforced average distance between the target customer and each cluster center.

[0202] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0203] For each cluster center, the distance between each sub-attribute information and its corresponding sub-cluster center is determined based on each sub-attribute information in the target attribute information and the corresponding sub-cluster center.

[0204] Based on the distance between each sub-attribute information and the corresponding sub-cluster center, the ordered reinforced average distance between the target customer and the cluster center is determined.

[0205] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0206] Determine the absolute value of the first difference between each sub-attribute information and its corresponding sub-cluster center;

[0207] For each sub-attribute information, determine the maximum value from the sub-attribute information and the corresponding sub-cluster centers;

[0208] The distance between the sub-attribute information and the corresponding sub-cluster center is determined by the ratio of the absolute value to the maximum value.

[0209] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0210] The ratio is used as the distance between the sub-attribute information and the corresponding sub-cluster center.

[0211] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0212] Determine the product between each distance and the corresponding target weight; the target weight is determined based on the target ordered augmentation average model, which is obtained by training a preset initial ordered augmentation average model on a training set corresponding to a preset customer type;

[0213] The summation result is obtained by summing the products of each distance and its corresponding target weight.

[0214] The summation result is used as the ordered reinforced average distance between the target customer and the cluster center.

[0215] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0216] The preset customer type corresponding to the target cluster center is taken as the customer type of the target customer; the target cluster center is the cluster center corresponding to the smallest ordered reinforced average distance.

[0217] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0218] Obtain the training set corresponding to the preset customer type;

[0219] Based on the training set corresponding to the preset customer type and the preset customer type, the initial ordered augmented average model is trained to obtain the target ordered augmented average model.

[0220] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0221] Obtain attribute information samples of each customer in the training set corresponding to each preset customer type.

[0222] For each preset customer type, the sub-cluster centers corresponding to each preset customer type are determined based on the attribute information samples of the same type of customers in the training set corresponding to the preset customer type.

[0223] Based on the sub-cluster centers corresponding to the preset customer types, determine the cluster centers corresponding to the preset customer types.

[0224] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0225] Obtain target attribute information of target customers;

[0226] Based on the target attribute information and the cluster centers corresponding to each preset customer type, determine the ordered reinforced average distance between the target customer and each cluster center;

[0227] The customer type of the target customer is determined based on the ordered reinforced average distance between the target customer and each cluster center.

[0228] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0229] For each cluster center, the distance between each sub-attribute information and its corresponding sub-cluster center is determined based on each sub-attribute information in the target attribute information and the corresponding sub-cluster center.

[0230] Based on the distance between each sub-attribute information and the corresponding sub-cluster center, the ordered reinforced average distance between the target customer and the cluster center is determined.

[0231] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0232] Determine the absolute value of the first difference between each sub-attribute information and its corresponding sub-cluster center;

[0233] For each sub-attribute information, determine the maximum value from the sub-attribute information and the corresponding sub-cluster centers;

[0234] The distance between the sub-attribute information and the corresponding sub-cluster center is determined by the ratio of the absolute value to the maximum value.

[0235] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0236] The ratio is used as the distance between the sub-attribute information and the corresponding sub-cluster center.

[0237] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0238] Determine the product between each distance and the corresponding target weight; the target weight is determined based on the target ordered augmentation average model, which is obtained by training a preset initial ordered augmentation average model on a training set corresponding to a preset customer type;

[0239] The summation result is obtained by summing the products of each distance and its corresponding target weight.

[0240] The summation result is used as the ordered reinforced average distance between the target customer and the cluster center.

[0241] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0242] The preset customer type corresponding to the target cluster center is taken as the customer type of the target customer; the target cluster center is the cluster center corresponding to the smallest ordered reinforced average distance.

[0243] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0244] Obtain the training set corresponding to the preset customer type;

[0245] Based on the training set corresponding to the preset customer type and the preset customer type, the initial ordered augmented average model is trained to obtain the target ordered augmented average model.

[0246] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0247] Obtain attribute information samples of each customer in the training set corresponding to each preset customer type.

[0248] For each preset customer type, the sub-cluster centers corresponding to each preset customer type are determined based on the attribute information samples of the same type of customers in the training set corresponding to the preset customer type.

[0249] Based on the sub-cluster centers corresponding to the preset customer types, determine the cluster centers corresponding to the preset customer types.

[0250] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0251] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0252] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0253] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method of classifying customers, characterized by, The method comprises: obtaining target attribute information of a target customer; determining an ordered reinforced average distance between the target customer and each cluster center corresponding to each preset customer type according to the target attribute information and the cluster center; determining a customer type of the target customer according to the ordered reinforced average distance between the target customer and each cluster center; The method comprises: for each cluster center, determining a distance between each sub attribute information in the target attribute information and a corresponding sub cluster center according to the sub attribute information and the sub cluster center; determining the ordered reinforced average distance between the target customer and the cluster center according to the distance between each sub attribute information and the corresponding sub cluster center; The method comprises: determining an absolute value of a first difference between each sub attribute information and the corresponding sub cluster center; for each sub attribute information, determining a maximum value from the sub attribute information and the corresponding sub cluster center; taking a ratio of the absolute value and the maximum value as the distance between the sub attribute information and the corresponding sub cluster center; The method comprises: determining a product result between each distance and a corresponding target weight; the target weight is determined according to a target ordered reinforced average model, which is obtained by training an initial ordered reinforced average model based on a training set corresponding to the preset customer type; summing the product results between each distance and the corresponding target weight to obtain a summation result; taking the summation result as the ordered reinforced average distance between the target customer and the cluster center.

2. The method of claim 1, wherein, The method comprises: taking a preset customer type corresponding to a target cluster center as the customer type of the target customer; the target cluster center is a cluster center corresponding to the smallest ordered reinforced average distance.

3. The method of claim 2, wherein, The method further comprises: obtaining a training set corresponding to the preset customer type; training the initial ordered reinforced average model according to the training set corresponding to the preset customer type and the preset customer type to obtain a target ordered reinforced average model.

4. The method of claim 3, wherein, The method further comprises: obtaining attribute information samples of each customer in each training set corresponding to each preset customer type; for each preset customer type, determining each sub cluster center corresponding to the preset customer type according to attribute information samples of customers of the same type in the training set corresponding to the preset customer type; determining a cluster center corresponding to the preset customer type according to each sub cluster center corresponding to the preset customer type.

5. A client classification apparatus characterized by comprising: The device comprises: The first obtaining module is configured to obtain target attribute information of a target customer; The first determining module is configured to determine an ordered reinforced average distance between the target customer and each cluster center according to the target attribute information and the cluster center corresponding to each preset customer type; The second determining module is configured to determine a customer type of the target customer according to the ordered reinforced average distance between the target customer and each cluster center; The first determining module is specifically configured to: For each cluster center, determine an absolute value of a first difference value between each sub attribute information in the target attribute information and a corresponding sub cluster center; For each sub attribute information, determine a maximum value from the sub attribute information and the corresponding sub cluster center; Take a ratio of the absolute value and the maximum value as a distance between the sub attribute information and the corresponding sub cluster center; Determine the ordered reinforced average distance between the target customer and the cluster center according to the distance between each sub attribute information and the corresponding sub cluster center; The determination of the ordered reinforced average distance between the target customer and the cluster center according to the distance between each sub attribute information and the corresponding sub cluster center includes: Determine a product result between each distance and a corresponding target weight; the target weight is determined according to a target ordered reinforced average model, which is obtained by training an initial ordered reinforced average model based on a training set corresponding to the preset customer type; Sum the product results between each distance and the corresponding target weight to obtain a summation result; Take the summation result as the ordered reinforced average distance between the target customer and the cluster center. 6.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-5 when the computer program is executed by the processor. The processor executes the computer program to implement the steps of the method in any one of claims 1 to 4.

7. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 4.

8. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 4.

Images