CN118245341A

CN118245341A - Service model switching method, device, electronic equipment and computer readable medium

Info

Publication number: CN118245341A
Application number: CN202410230121.9A
Authority: CN
Inventors: 赵子龙; 吴谦
Original assignee: Park Road Credit Information Co ltd
Current assignee: Park Road Credit Information Co ltd
Filing date: 2024-02-29
Publication date: 2024-06-25

Abstract

The embodiment of the disclosure discloses a service model switching method, a service model switching device, an electronic device and a computer readable medium. One embodiment of the method comprises the following steps: performing performance detection on the user early warning information generation model to be switched and the target user early warning information generation model; determining whether a user early warning information generation model to be switched meets a preset quick switching condition or not; determining a preset result weight information group in the preset result weight information group sequence as a target result weight information group; inputting the business data information into a user early warning information generation model to be switched and a target user early warning information generation model; the first output result information and the second output result information are combined. According to the embodiment, the situation that the service flow is interrupted due to the fact that the original model and the new model are directly switched is reduced, the experience of service users is improved, the possibility of leakage of service data is reduced, and the situation that the stability of a service system is poor due to the fact that the new model and the old model are switched is reduced.

Description

Service model switching method, device, electronic equipment and computer readable medium

Technical Field

Embodiments of the present disclosure relate to the field of computer technologies, and in particular, to a method and an apparatus for switching a service model, an electronic device, and a computer readable medium.

Background

The updating of the service model can be better adapted to the updated service data. At present, when iterating a user early warning information generation model, the following modes are generally adopted: first, training the latest user early warning information generation model according to the latest business data information. Then, the original user early warning information generation model is directly switched into the latest trained user early warning information generation model, so that the latest user early warning information generation model can be used for processing the business data information.

However, when the user early warning information generation model is switched in the above manner, there are often the following technical problems:

The original user early warning information generation model is directly switched to the latest user early warning information generation model, the original user early warning information generation model and the latest user early warning information generation model are different, the situation that the business process is easily interrupted is directly switched between the original model and the new model, the business system is easily caused to malfunction, the experience of the business user is poor, the possibility of leakage of the business data is high, and the stability of the business system is easily caused by directly using the new model.

The above information disclosed in this background section is only for enhancement of understanding of the background of the inventive concept and, therefore, may contain information that does not form the prior art that is already known to those of ordinary skill in the art in this country.

Disclosure of Invention

The disclosure is in part intended to introduce concepts in a simplified form that are further described below in the detailed description. The disclosure is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

Some embodiments of the present disclosure propose a method, apparatus, electronic device, and computer-readable medium for switching a service model to solve one or more of the technical problems mentioned in the background section above.

In a first aspect, some embodiments of the present disclosure provide a method for switching a service model, where the method includes: performing performance detection processing on a user early warning information generation model to be switched and a target user early warning information generation model to obtain service model performance information corresponding to the user early warning information generation model to be switched and service model performance information corresponding to the target user early warning information generation model, wherein the target user early warning information generation model is used for replacing the user early warning information generation model to be switched; determining whether the early warning information generation model of the user to be switched meets a preset quick switching condition according to the service model performance information of the early warning information generation model of the user to be switched; in response to determining that the user early warning information generation model to be switched does not meet the preset quick switching condition, determining whether preset switching requirement information corresponding to the user early warning information generation model to be switched meets the preset quick switching requirement condition; in response to determining that the preset switching requirement information corresponding to the user early warning information generation model to be switched does not meet the preset quick switching requirement condition, determining a preset result weight information group meeting the preset weight information sequence condition in a preset result weight information group sequence as a target result weight information group; according to the preset result weight information group sequence and the target result weight information group, for the user early warning information generation model to be switched, executing the following switching steps: deleting the target result weight information group from the preset result weight information group sequence to update the preset result weight information group sequence; in response to detecting input request information of service data information, and the target result weight information set meets a preset result weight condition, inputting the service data information into the to-be-switched user early warning information generation model and the target user early warning information generation model to obtain output result information of the to-be-switched user early warning information generation model as first output result information and output result information of the target user early warning information generation model as second output result information; and according to the target result weight information group, carrying out combination processing on the first output result information and the second output result information to obtain service data output result information corresponding to the service data information, so as to carry out switching processing on the early warning information generation model of the user to be switched.

In a second aspect, some embodiments of the present disclosure provide a switching device for a service model, including a detection unit configured to perform performance detection processing on a to-be-switched user early-warning information generation model and a target user early-warning information generation model, to obtain service model performance information corresponding to the to-be-switched user early-warning information generation model and service model performance information corresponding to the target user early-warning information generation model, where the target user early-warning information generation model is used to replace the to-be-switched user early-warning information generation model; the first determining unit is configured to determine whether the user early warning information generating model to be switched meets a preset quick switching condition according to the service model performance information of the user early warning information generating model to be switched; a second determining unit configured to determine whether preset switching requirement information corresponding to the user early warning information generating model to be switched meets a preset quick switching requirement condition in response to determining that the user early warning information generating model to be switched does not meet the preset quick switching condition; a third determining unit configured to determine, as a target result weight information group, a preset result weight information group satisfying a preset weight information order condition in a preset result weight information group sequence in response to determining that preset switching requirement information corresponding to the user early warning information generation model to be switched does not satisfy the preset fast switching requirement condition; the switching unit is configured to execute the following switching steps for the user early warning information generation model to be switched according to the preset result weight information group sequence and the target result weight information group: deleting the target result weight information group from the preset result weight information group sequence to update the preset result weight information group sequence; in response to detecting input request information of service data information, and the target result weight information set meets a preset result weight condition, inputting the service data information into the to-be-switched user early warning information generation model and the target user early warning information generation model to obtain output result information of the to-be-switched user early warning information generation model as first output result information and output result information of the target user early warning information generation model as second output result information; and according to the target result weight information group, carrying out combination processing on the first output result information and the second output result information to obtain service data output result information corresponding to the service data information, so as to carry out switching processing on the early warning information generation model of the user to be switched.

In a third aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device having one or more programs stored thereon, which when executed by one or more processors causes the one or more processors to implement the method described in any of the implementations of the first aspect above.

In a fourth aspect, some embodiments of the present disclosure provide a computer readable medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method described in any of the implementations of the first aspect.

The above embodiments of the present disclosure have the following advantageous effects: according to the service model switching method, the condition of service flow interruption can be reduced, the experience of service users is improved, the possibility of leakage of service data is reduced, and the stability of a service system is improved. Specifically, the situation that the service flow is easily interrupted is caused, the experience of the service user is poor, the possibility of service data leakage is high, and the reason that the stability of the service system is poor is caused by directly using the new model is that: the original user early warning information generation model is directly switched to the latest user early warning information generation model, the original user early warning information generation model and the latest user early warning information generation model are different, the situation that the business process is interrupted easily occurs when the original model and the new model are directly switched, the business system is easy to fail, the experience of the business user is poor and the possibility of leakage of the business data is high, And the direct use of the new model is liable to cause poor stability of the service system. Based on this, in the method for switching the service model according to some embodiments of the present disclosure, first, performance detection processing is performed on a to-be-switched user early-warning information generation model and a target user early-warning information generation model to obtain service model performance information corresponding to the to-be-switched user early-warning information generation model and service model performance information corresponding to the target user early-warning information generation model, where the target user early-warning information generation model is used to replace the to-be-switched user early-warning information generation model. Therefore, the performance information of the user early warning information generation model to be switched and the target user early warning information generation model can be obtained, and the method can be used for determining the mode of switching the user early warning information generation model to be switched. And then, determining whether the early warning information generation model of the user to be switched meets the preset quick switching condition according to the service model performance information of the early warning information generation model of the user to be switched. Therefore, whether the requirement for fast switching of the user early warning information generation model to be switched exists can be determined. And then, in response to determining that the user early warning information generation model to be switched does not meet the preset quick switching condition, determining whether preset switching requirement information corresponding to the user early warning information generation model to be switched meets the preset quick switching requirement condition. Therefore, whether the preset switching requirement information corresponding to the user early warning information generation model to be switched is the requirement of quick switching can be determined. and then, in response to determining that the preset switching requirement information corresponding to the user early warning information generation model to be switched does not meet the preset quick switching requirement condition, determining a preset result weight information group meeting the preset weight information sequence condition in a preset result weight information group sequence as a target result weight information group. Therefore, the target result weight information group can be selected to switch the user early warning information generation model to be switched. Then, according to the preset result weight information group sequence and the target result weight information group, for the user early warning information generation model to be switched, the following switching steps are executed: first, the target result weight information group is deleted from the preset result weight information group sequence to update the preset result weight information group sequence. therefore, the preset result weight information group sequence can be updated, and the user early warning information generation model to be switched can be switched according to the updated preset result weight information group sequence. And then, in response to detecting the input request information of the service data information, and the target result weight information set meets the preset result weight condition, inputting the service data information into the to-be-switched user early warning information generation model and the target user early warning information generation model to obtain the output result information of the to-be-switched user early warning information generation model as first output result information and the output result information of the target user early warning information generation model as second output result information. Thus, output result information corresponding to the service data information can be obtained. And finally, according to the target result weight information group, combining the first output result information and the second output result information to obtain service data output result information corresponding to the service data information, so as to perform switching processing on the user early warning information generation model to be switched. Therefore, the output result information of the user early warning information generation model to be switched and the output result information of the target user early warning information generation model can be combined through the selected target result weight information group. And because the user early warning information generation model to be switched is not directly switched to the target user early warning information generation model, but is gradually switched to the target user early warning information generation model by continuously selecting the target result weight information group. therefore, the situation that the service flow is interrupted due to direct switching of the original model and the new model is reduced, the experience of service users is improved, the possibility of leakage of service data is reduced, and the situation that the stability of a service system is poor due to direct switching of the new model and the old model is reduced.

Drawings

The above and other features, advantages, and aspects of embodiments of the present disclosure will become more apparent by reference to the following detailed description when taken in conjunction with the accompanying drawings. The same or similar reference numbers will be used throughout the drawings to refer to the same or like elements. It should be understood that the figures are schematic and that elements and components are not necessarily drawn to scale.

FIG. 1 is a flow chart of some embodiments of a handoff method of a business model according to the present disclosure;

FIG. 2 is a schematic structural diagram of some embodiments of a switching apparatus of a business model according to the present disclosure;

Fig. 3 is a schematic structural diagram of an electronic device suitable for use in implementing some embodiments of the present disclosure.

Detailed Description

Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While certain embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete. It should be understood that the drawings and embodiments of the present disclosure are for illustration purposes only and are not intended to limit the scope of the present disclosure.

It should be noted that, for convenience of description, only the portions related to the present invention are shown in the drawings. Embodiments of the present disclosure and features of embodiments may be combined with each other without conflict.

It should be noted that the terms "first," "second," and the like in this disclosure are merely used to distinguish between different devices, modules, or units and are not used to define an order or interdependence of functions performed by the devices, modules, or units.

It should be noted that references to "one", "a plurality" and "a plurality" in this disclosure are intended to be illustrative rather than limiting, and those of ordinary skill in the art will appreciate that "one or more" is intended to be understood as "one or more" unless the context clearly indicates otherwise.

The names of messages or information interacted between the various devices in the embodiments of the present disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

The present disclosure will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.

Fig. 1 illustrates a flow 100 of some embodiments of a handoff method for a business model according to the present disclosure. The switching method of the service model comprises the following steps:

And step 101, performing performance detection processing on the early warning information generation model of the user to be switched and the early warning information generation model of the target user to obtain service model performance information corresponding to the early warning information generation model of the user to be switched and service model performance information corresponding to the early warning information generation model of the target user.

In some embodiments, the execution body of the service model switching method may perform performance detection processing on the to-be-switched user early-warning information generation model and the target user early-warning information generation model to obtain service model performance information corresponding to the to-be-switched user early-warning information generation model and service model performance information corresponding to the target user early-warning information generation model. The target user early warning information generation model can be used for replacing the user early warning information generation model to be switched. The user early warning information generation model to be switched and the target user early warning information generation model are both models for generating early warning information of corresponding service users. The early warning information can represent the default rate of the service user. The business user may be a borrower. The default rate may be a probability that the target user cannot repay the borrow in accordance with a repayment agreement. For example, the user early warning information generation model to be switched and the target user early warning information generation model can be feedforward neural network models. The service model performance information of the user early warning information generation model to be switched can represent the accuracy rate, recall rate and precision rate of the user early warning information generation model to be switched. The service model performance information of the target user early warning information generation model can represent the accuracy rate, recall rate and precision rate of the target user early warning information generation model. In practice, the execution body may perform performance detection processing on the to-be-switched user early-warning information generation model and the target user early-warning information generation model according to various modes, so as to obtain service model performance information corresponding to the to-be-switched user early-warning information generation model and service model performance information corresponding to the target user early-warning information generation model.

In some optional implementations of some embodiments, the executing body may perform performance detection processing on the to-be-switched user early-warning information generation model and the target user early-warning information generation model to obtain service model performance information corresponding to the to-be-switched user early-warning information generation model and service model performance information corresponding to the target user early-warning information generation model by:

First, a historical business data information set is obtained. The historical service data information in the historical service data information set can comprise basic information of service users, value related information of the service users, historical credibility information of the service users and market circulation environment information. The historical credibility information of the service user can be the default rate of the service user of the history record. The basic information of the service user can include an identification card number of the service user, a name of the service user, an address of the service user, and a occupation of the service user. The value related information of the service user can represent the value condition of the service user. For example, the value related information may be the property status of business users (car, house, wage status). The market circulation environment information can represent domestic production total values. In practice, the executing entity may obtain the set of historical service data information from a database storing historical service data information.

And secondly, performing data cleaning processing on each history service data information in the acquired history service data information set to obtain a history service data information set after the data cleaning processing as a target history service data information set. The target historical service data information set can represent the historical service data information set after data cleaning processing. In practice, the execution body may perform data cleaning processing on each history service data information in the acquired history service data information set in various manners, so as to obtain a history service data information set after the data cleaning processing as a target history service data information set.

And thirdly, respectively inputting each target historical service data information in the target historical service data information set into the to-be-switched user early warning information generation model and the target user early warning information generation model.

And fourthly, determining each piece of prediction result information corresponding to the target historical business data information set output by the user early warning information generation model to be switched as each piece of first prediction result information. Wherein the first prediction result information in the first prediction result information corresponds to the target historical service data information in the target historical service data information set. The first prediction result information in the first prediction result information may be the default rate of the corresponding target historical service data information output by the early warning information generation model of the user to be switched. For example, the first predictor information may be 40%.

And fifthly, determining each piece of prediction result information corresponding to the target historical business data information set output by the target user early warning information generation model as each piece of second prediction result information. Wherein the second prediction result information in the second prediction result information corresponds to the target historical service data information in the target historical service data information set. The second prediction result information in the second prediction result information may be the default rate of the corresponding target historical service data information output by the target user early warning information generation model. For example, the second predictor information may be 30%.

And sixthly, generating service model performance information corresponding to the user early warning information generation model to be switched and service model performance information corresponding to the target user early warning information generation model according to the actual early warning result information set corresponding to the target historical service data information set, the first prediction result information and the second prediction result information. The actual early warning result information in the actual early warning result information set can represent the actual default rate of the corresponding target historical service data information. For example, the actual early warning result information may be: "target history business data information: xx; actual breach rate: 60% ".

However, in practice, it has been found that when testing the performance of a model according to the historical business data information set through the present disclosure, there are often the following technical problems:

The method comprises the steps that a historical service data information set is directly input into a user early warning information generation model to be switched and a target user early warning information generation model, preprocessing is not considered for the historical service data information set, and when more redundant data or more error data exist in the historical service data information set, the accuracy of results output by the user early warning information generation model to be switched and the target user early warning information generation model is poor, and the accuracy of performance of the user early warning information generation model to be switched and the target user early warning information generation model to be switched is low.

In view of the above technical problems, the following solutions are decided to be adopted:

in some optional implementations of some embodiments, the executing body may perform data cleaning processing on each of the acquired historical service data information in the obtained historical service data information set by using the historical service data information set after the data cleaning processing as the target historical service data information set:

the first step, mapping the historical service data information set to obtain a historical service data information vector sequence corresponding to the historical service data information set. The length information corresponding to the historical service data information vector sequence may be preset window length information. The preset window length information may be a preset length of a time-series data segment for constructing the track matrix. For example, the preset window length information may be 3. The above-mentioned historical service data information vector sequence may be a set of historical service data information after mapping processing. The specific value of the preset window length information is not specifically limited herein. In practice, the execution main body can map the historical service data information set by a singular spectrum analysis method to obtain a historical service data information vector sequence.

And secondly, sorting the historical service data information sets according to the preset window length information to obtain a historical service data information matrix corresponding to the historical service data information sets. The history service data information matrix may be a track matrix corresponding to the history service data information. In practice, the execution body may perform the sorting process on the historical service data information set in a hysteresis arrangement manner, so as to obtain a historical service data information matrix corresponding to the historical service data information set.

And thirdly, decomposing the historical service data information matrix to obtain a decomposed historical service data information matrix serving as a target historical service data information matrix. The target historical service data information matrix can represent the historical service data information matrix after decomposition processing. In practice, the execution body may decompose the historical service data information matrix by using a singular value decomposition method.

And step four, dividing the index information set corresponding to the target historical service data information matrix to obtain the index information set after dividing. The subscript information in the subscript information set may represent a subscript of the target historical data information in the target historical service data information matrix. For example, the set of subscript information may include 1,2, 3. The subscript information in the subscript information set may correspond to the target historical service data information in the target historical service data information matrix. In practice, the execution body may combine each preset number of subscript information in the subscript information set into a subscript information set, to obtain each subscript information set as a subscript information set after the division processing. Here, specific values of the preset values are not particularly limited. For example, the preset value may be 5.

And fifthly, carrying out grouping processing on the target historical service data information matrix according to the index information set after the division processing. The divided subscript information sets may represent each subscript information group. For example, the set of subscript information may include 1,2,3, 4. In practice, for each index information group in the index information set after the division processing, the execution body may determine each target history service data information corresponding to the index information group in the target history service data information matrix as a target history service data information group, and obtain each target history service data information group, so as to perform packet processing on the target history service data information matrix.

And sixthly, performing conversion processing on the target historical service data information matrix after the grouping processing to obtain the target historical service data information matrix after the conversion processing. The target historical service data information matrix after the conversion processing can represent a one-dimensional matrix. In practice, the execution body may perform conversion processing on the target historical service data information matrix after the packet processing in a diagonally averaged manner. Thus, the two-dimensional target historical service data information matrix after grouping processing can be converted into a one-dimensional target historical service data information matrix.

Seventh, the history service data information set is subjected to data reconstruction processing, and the history service data information set after the data reconstruction processing is obtained and is used as a reconstructed history service data information set. The reconstructed historical service data information may represent a set of the reconstructed historical service data information. In practice, first, the executing body may perform descending order sorting processing on the historical service data information set according to the singular value corresponding to each historical service data information in the historical service data information set, to obtain a historical service data information set after the descending order sorting processing. Then, for each of the history service data information sets after the descending order sorting process, the execution body may determine a square value corresponding to each of the history service data information sets. And then, determining the square value of the singular value corresponding to the historical service data information as a target square value. And secondly, determining the ratio between the obtained target square value and the sum of the obtained square values as the contribution rate corresponding to the historical service data information. And finally, determining the previous t component of each history service data information with the contribution rate larger than the preset contribution rate in the history service data information set to reconstruct the history service data information set. Here, the specific value of the preset contribution ratio is not specifically limited. For example, the preset contribution rate may be 85%. Here, the specific value of the first t component is not particularly limited.

Eighth, generating residual sequence information corresponding to the reconstructed historical service data information set and the historical service data information set according to the reconstructed historical service data information set and the historical service data information set. Wherein, the residual sequence information may characterize the residual sequence. The history service data information in the history service data information set corresponds to the reconstructed history service data information in the reconstructed history service data information set. In practice, for each history service data information in the history service data information set, the executing body may determine a difference between the history service data information and reconstructed history service data information corresponding to the history service data information in the reconstructed history service data information set as a target difference. Then, each obtained target difference value is determined as residual sequence information.

And a ninth step of generating sequence mean value information and mean square error information corresponding to the residual sequence information according to the residual sequence information. The sequence average information may represent an average value of the residual sequence information. The mean square error information may characterize a mean square error of the residual sequence information. In practice, first, the execution subject may determine the average value of the residual sequence information as the sequence average value information corresponding to the residual sequence information. Then, the mean square error of the residual sequence information is determined as the mean square error information corresponding to the residual sequence information.

And tenth, performing data cleaning processing on the residual sequence information according to the sequence mean value information and the mean square error information so as to perform data cleaning processing on the acquired historical service data information set. In practice, for each residual value in the residual sequence information, the execution body may determine an absolute value of a difference between the residual value and an average value of the residual sequence information as a target absolute value. Then, the ratio of the target absolute value to the mean square error is determined as a target ratio. And finally, deleting the historical service data information corresponding to the residual value corresponding to the target ratio from the historical service data information set in response to the fact that the target ratio is larger than a preset value, so as to perform data cleaning processing on the obtained historical service data information set.

The technical scheme is taken as an invention point of the embodiment of the disclosure, and solves the technical problems: the method comprises the steps that a historical service data information set is directly input into a user early warning information generation model to be switched and a target user early warning information generation model, preprocessing is not considered for the historical service data information set, and when more redundant data or more error data exist in the historical service data information set, the accuracy of results output by the user early warning information generation model to be switched and the target user early warning information generation model is poor, and the accuracy of performance of the user early warning information generation model to be switched and the target user early warning information generation model to be switched is low. The accuracy of the results output by the to-be-switched user early warning information generation model and the target user early warning information generation model is low, and the factors of the low accuracy of the model performances of the predicted to-be-switched user early warning information generation model and the target user early warning information generation model are often as follows: and directly inputting the historical service data information set into the to-be-switched user early warning information generation model and the target user early warning information generation model, wherein preprocessing of the historical service data information set is not considered. If the factors are solved, the effect of improving the accuracy of the results output by the to-be-switched user early warning information generation model and the target user early warning information generation model can be achieved, and the accuracy of the performances of the predicted to-be-switched user early warning information generation model and the target user early warning information generation model can be improved.

In some optional implementations of some embodiments, the executing entity may generate the service model performance information corresponding to the user early warning information generation model to be switched and the service model performance information corresponding to the target user early warning information generation model according to an actual early warning result information set corresponding to the target historical service data information set, the respective first prediction result information, and the respective second prediction result information by:

And determining each piece of first prediction result information output by the user early warning information generation model to be switched as each piece of target prediction result information. The target prediction result information corresponding to the user early warning information generation model to be switched can represent the prediction result information output by the user early warning information generation model to be switched.

Second, for each target prediction result information in the respective target prediction result information, the following detection steps are performed:

and a first detection step of determining target historical service data information corresponding to the target prediction result information in the target historical service data information set as historical service data information.

And a second detection step, determining the actual early warning result information corresponding to the historical service data information in the actual early warning result information set as target actual early warning result information. The actual early warning result information in the actual early warning result information set corresponds to the target historical service data information in the target historical service data information set. The target actual early warning result information can represent the actual default rate corresponding to the historical service data information. In practice, the executing body may determine, as the target actual warning result information corresponding to the historical service data information, the actual warning result information, which is the same as the target historical service data information and is included in the actual warning result information set.

And a third detection step, namely performing similarity detection processing on the determined target actual early warning result information and the target prediction result information to obtain similarity information corresponding to the determined target actual early warning result information and the target prediction result information. The similarity information may represent a similarity between the determined target actual early warning result information and the target prediction result information. In practice, first, the executing body may determine, as the target difference, a difference between the breach rate represented by the target actual warning result information and the breach rate represented by the target prediction result information. Then, it is determined whether the target difference is greater than a preset default rate difference. Here, the specific value of the preset default rate difference is not specifically limited herein. And then, in response to determining that the target difference value is greater than the preset default rate difference value, determining the first preset similarity as similarity information corresponding to the determined target actual early warning result information and the target prediction result information. Here, the specific value of the first preset similarity is not specifically limited herein. For example, the first preset similarity may be 20%. And secondly, in response to determining that the target difference value is smaller than or equal to the preset default rate difference value, determining second preset similarity as similarity information corresponding to the determined target actual early warning result information and the target prediction result information. Here, the specific value of the second preset similarity is not specifically limited herein. For example, the second preset similarity may be 80%.

And thirdly, taking each piece of second predicted result information output by the target user early warning information generation model as each piece of target predicted result information, and executing the detection step again for each piece of target predicted result information in each piece of target predicted result information to obtain each piece of similarity information corresponding to the target user early warning information generation model. The similarity information in each piece of similarity information corresponding to the target user early warning information generation model may represent similarity between the second prediction result information output by the target user early warning information generation model and the corresponding target actual early warning result information.

And fourthly, generating service model performance information corresponding to the user early warning information generation model to be switched and service model performance information corresponding to the target user early warning information generation model according to the similarity information corresponding to the user early warning information generation model to be switched, the similarity information corresponding to the target user early warning information generation model and the actual early warning result information set. In practice, the executing body may generate, in various manners, service model performance information corresponding to the user early warning information generation model to be switched and service model performance information corresponding to the target user early warning information generation model according to each piece of similarity information corresponding to the user early warning information generation model to be switched, each piece of similarity information corresponding to the target user early warning information generation model, and the actual early warning result information set.

However, in practice, it is found that when a new and an old model are switched in the manner of the present disclosure, there are often the following technical problems:

And switching the user early warning information generation model to be switched according to the target user early warning information generation model directly according to the switching requirement, wherein the switching processing of the user early warning information generation model to be switched according to the model performance of the target user early warning information generation model and the user early warning information generation model to be switched is not considered, and if a new model with poor performance is used, the accuracy of a new model prediction result is lower.

In some optional implementations of some embodiments, the executing body may generate the service model performance information corresponding to the user early warning information generation model to be switched and the service model performance information corresponding to the target user early warning information generation model according to each similarity information corresponding to the user early warning information generation model to be switched, each similarity information corresponding to the target user early warning information generation model, and the actual early warning result information set by:

And determining each piece of similarity information corresponding to the user early warning information generation model to be switched as each piece of target similarity information. The target similarity information in the target similarity information may represent similarity between the predicted result information and the actual early-warning result information output by the early-warning information generation model of the user to be switched.

Secondly, according to the actual early warning result information set, for each target similarity information in each target similarity information, executing the following early warning steps:

And a first early warning step, in response to determining that the target similarity information meets a preset similarity condition and the target prediction result information corresponding to the target similarity information meets a preset early warning information condition, determining a preset prediction result category information group which characterizes a category to be early-warning processed in a preset prediction result category information group set as a target preset prediction result category information group corresponding to the target similarity information. The preset early warning information condition may be that the default ratio represented by the target prediction result information is greater than or equal to a preset default ratio value. Here, the specific value of the preset default ratio value is not particularly limited. For example, the preset violation rate value may be 70%. The set of preset prediction result category information sets may include a preset number of preset prediction result category information sets. The predetermined number may be 4. The preset prediction result category information set may include a preset prediction result category information set for representing a category to be pre-warned, wherein the pre-warned and actual pre-warned prediction result category information set for representing a category to be pre-warned and the pre-warned prediction result category information set for representing a category to be pre-warned. The preset similarity condition may be that the similarity represented by the target similarity information is equal to the second preset similarity. The preset predicted result category information group for representing the category to be pre-warned may include corresponding actual pre-warning result information and corresponding similarity information of which the default rate represented by the target predicted result information is greater than or equal to a preset default rate value.

And a second early warning step of determining a preset prediction result category information group representing a category not requiring early warning processing in the preset prediction result category information group set as a target preset prediction result category information group corresponding to the target similarity information in response to determining that the target similarity information meets the preset similarity condition and the target prediction result information corresponding to the target similarity information does not meet the preset early warning information condition. The preset predicted result category information group for representing the no-early-warning processing category can comprise corresponding actual early-warning result information and corresponding default rate of target predicted result information representation which are smaller than each similarity information of the preset default rate value.

And a third pre-warning step of determining, as a target pre-prediction result category information set corresponding to the target similarity information, a pre-prediction result category information set in which the pre-warning process is not required for the characterization prediction in the pre-prediction result category information set and the pre-warning process is actually required, in response to determining that the target similarity information does not satisfy the pre-preset similarity condition and the target prediction result information corresponding to the target similarity information does not satisfy the pre-warning information condition. The preset predicted result category information group of the predicted category which does not need early warning processing and needs actual early warning processing can contain various similarity information that the corresponding actual early warning result information represents that the default ratio is larger than the preset default ratio value and the corresponding target predicted result information represents that the default ratio is smaller than or equal to the preset default ratio value.

And a fourth early warning step of determining, as a target preset predicted result category information set corresponding to the target similarity information, a preset predicted result category information set which characterizes a predicted type requiring early warning processing and does not actually require early warning processing in the set of preset predicted result category information sets in response to determining that the target similarity information does not satisfy the preset similarity condition and that target predicted result information corresponding to the target similarity information satisfies the preset early warning information condition. The preset predicted result category information group of the predicted type which is characterized by requiring early warning processing and does not require early warning processing in practice can contain various similarity information of which the corresponding default rate of the target predicted result information characterization is larger than the preset default rate value and the corresponding default rate of the actual early warning result information characterization is smaller than or equal to the preset default rate value.

And fifth early warning step, the quantity information of the target similarity information corresponding to the determined target preset prediction result category information group is updated, and the updated target preset prediction result category information group is obtained. The number information may represent the number of target similarity information represented by the target preset prediction result category information group. In practice, the execution body may add 1 to the number of the number information characterizations to obtain the updated target preset prediction result category information set.

Thirdly, building early warning matrix information corresponding to the early warning information generation model of the user to be switched according to the updated target preset prediction result category information groups. The early warning matrix information may be a matrix for early warning the early warning information generation model of the user to be switched. The early warning matrix information corresponding to the early warning information generation model of the user to be switched may be a matrix for generating performance of the early warning information generation model of the user to be switched. The pre-warning matrix information corresponding to the pre-warning matrix information of the pre-warning information generation model of the user to be switched may be a 2×2 two-dimensional matrix. In practice, first, the executing body may replace each preset predicted outcome category information set corresponding to the target predicted outcome category information set after the update processing in the preset predicted outcome category information set with the target preset predicted outcome category information set after the update processing, so as to update the preset predicted outcome category information set. And finally, randomly combining each preset prediction result category information group in the updated preset prediction result category information group set into a 2 multiplied by 2 matrix corresponding to the user early warning information generation model to be switched.

And fourthly, taking each piece of similarity information corresponding to the target user early-warning information generation model as each piece of target similarity information, and executing the early-warning step again for each piece of target similarity information in each piece of target similarity information according to the actual early-warning result information set to obtain each target preset prediction result category information group after updating processing of the target user early-warning information generation model.

Fifthly, generating early warning matrix information corresponding to the target user early warning information generation model according to each target preset prediction result category information group after updating processing of the target user early warning information generation model. The early warning matrix information corresponding to the target user early warning information generation model may be a matrix for generating performance of the target user early warning information generation model. The pre-warning matrix information corresponding to the pre-warning matrix information of the target user pre-warning information generation model may be a2×2 two-dimensional matrix. In practice, first, the executing body may replace each preset predicted outcome category information set corresponding to each target preset predicted outcome category information set in the preset predicted outcome category information set with each updated target preset predicted outcome category information set, so as to update the preset predicted outcome category information set. And finally, randomly combining each preset prediction result category information group in the updated preset prediction result category information group set into a2 multiplied by 2 matrix corresponding to the target user early warning information generation model.

And sixthly, generating service model performance information corresponding to the early warning information generation model of the user to be switched according to the early warning matrix information corresponding to the early warning information generation model of the user to be switched. In practice, first, the executing body may add the number of the pieces of similarity information included in the preset prediction result category information set for representing the category to be pre-warned in the pre-warning matrix information and the number of the pieces of similarity information included in the preset prediction result category information set for representing the category to be pre-warned, so as to obtain the first numerical value. And then, adding the quantity of similarity information representing the characteristics of the preset prediction result category information group of the category needing the early warning treatment, the quantity of each similarity information representing the characteristics of the preset prediction result category information group of the category needing no early warning treatment, the quantity of each similarity information representing the characteristics of the preset prediction result category information group which is required to be early warning treatment and does not actually need the early warning treatment and the quantity of each similarity information representing the characteristics of the preset prediction result category information group which is required to be early warning treatment and does not actually need the early warning treatment in the early warning matrix information to obtain a second numerical value. And then, determining the ratio of the first numerical value to the second numerical value as the accuracy of the early warning information generation model corresponding to the user to be switched. And secondly, adding the quantity of each piece of similarity information contained in the preset prediction result category information group representing the category to be pre-warned to the quantity of each piece of similarity information contained in the preset prediction result category information group representing the category to be pre-warned and not actually required to be pre-warned to obtain a third numerical value. And then, determining the ratio of the number of the similarity information contained in the preset prediction result category information group representing the category to be subjected to early-warning processing to the third numerical value as the accuracy rate of the early-warning information generation model corresponding to the user to be switched. And secondly, adding the quantity of the similarity information contained in the preset prediction result category information group of the category needing the early warning treatment and needing the early warning treatment in the characteristic prediction, so as to obtain a fourth numerical value. And then, determining the ratio between the quantity of each piece of similarity information contained in the preset prediction result category information group representing the category to be subjected to early-warning processing and the fourth numerical value as the recall rate of the early-warning information generation model corresponding to the user to be switched. And finally, determining the obtained accuracy, precision and recall rate as service model performance information of the early warning information generation model corresponding to the user to be switched.

And seventhly, generating service model performance information corresponding to the target user early warning information generation model according to the early warning matrix information corresponding to the target user early warning information generation model. In practice, the execution subject may generate the accuracy, precision and recall rate of the target user early warning information generation model according to the early warning matrix information corresponding to the target user early warning information generation model. It should be noted that, according to the pre-warning matrix information corresponding to the target user pre-warning information generation model, the manner of generating the service model performance information corresponding to the target user pre-warning information generation model is the same as the manner of generating the service model performance information corresponding to the user pre-warning information generation model to be switched according to the pre-warning matrix information corresponding to the user pre-warning information generation model to be switched. Therefore, the method for generating the service model performance information corresponding to the target user early warning information generation model is not described herein.

And eighth step, performing real-time display processing on the early warning matrix information of the early warning information generation model of the user to be switched and the early warning matrix information of the early warning information generation model of the target user. In practice, the obtained early warning matrix information of the early warning information generation model of the user to be switched and the early warning matrix information of the early warning information generation model of the target user are displayed on the associated display equipment in real time.

The technical scheme is taken as an invention point of the embodiment of the disclosure, and solves the technical problems: the user early warning information generation model to be switched is switched according to the target user early warning information generation model directly according to the switching requirement, the switching processing is not carried out on the user early warning information generation model to be switched according to the model performance of the target user early warning information generation model and the user early warning information generation model to be switched, and if a new model with poor performance is used, the accuracy of a new model prediction result is lower. Factors that lead to poor performance stability of the new model after switching and to low accuracy of the predicted results of the new model are often as follows: and directly switching the pre-warning information generation model of the user to be switched according to the switching requirement, and performing switching processing on the pre-warning information generation model of the user to be switched according to the model performance of the pre-warning information generation model of the target user and the pre-warning information generation model of the user to be switched. If the factors are solved, the performance stability of the new model after switching can be improved, and the accuracy of the predicted result of the new model is improved. Therefore, the performance stability of the new model after switching can be further improved, and the accuracy of the result predicted by the new model can be improved.

And 102, determining whether the early warning information generation model of the user to be switched meets the preset quick switching condition according to the service model performance information of the early warning information generation model of the user to be switched.

In some embodiments, the executing body may determine whether the early warning information generation model of the user to be switched meets a preset fast switching condition according to the service model performance information of the early warning information generation model of the user to be switched. The preset rapid switching condition can be that the accuracy, the precision and the recall rate of the service model performance information representation of the to-be-switched user early warning information generation model are smaller than or equal to preset values. Here, the specific value of the preset value is not particularly limited. For example, the preset value may be 60%. In practice, the executing body may determine whether the accuracy, recall rate and precision rate of the service model performance information representation of the switching user risk prediction service model are all greater than or equal to the preset values. And then, in response to determining that the accuracy rate, recall rate and precision rate of service model performance information representation of the switching user risk prediction service model are all larger than or equal to the preset numerical values, determining that the user early warning information generation model to be switched does not meet the preset rapid switching conditions. And then, in response to determining that the accuracy rate, recall rate and precision rate of service model performance information representation of the switching user risk prediction service model are smaller than the preset numerical values, determining that the user early warning information generation model to be switched meets the preset rapid switching condition.

And step 103, in response to determining that the pre-set fast switching condition is not satisfied by the pre-set switching requirement information corresponding to the pre-set fast switching requirement information generation model of the user to be switched, determining whether the pre-set fast switching requirement condition is satisfied by the pre-set switching requirement information corresponding to the pre-set fast switching requirement information generation model of the user to be switched.

In some embodiments, in response to determining that the user early-warning information generation model to be switched does not meet the preset fast switching condition, the execution body may determine whether preset switching requirement information corresponding to the user early-warning information generation model to be switched meets the preset fast switching requirement condition. The preset switching requirement information may characterize a mode of switching the user early warning information generation model to be switched. The preset switching requirement information may be a fast switching requirement or a smooth switching requirement. The preset fast switching requirement condition may be that the preset switching requirement information is a fast switching requirement.

And 104, determining a preset result weight information group meeting the preset weight information sequence condition in the preset result weight information group sequence as a target result weight information group in response to determining that the preset switching requirement information corresponding to the user early warning information generation model to be switched does not meet the preset quick switching requirement condition.

In some embodiments, in response to determining that the preset switching requirement information corresponding to the user early warning information generation model to be switched does not meet the preset fast switching requirement condition, the execution body may determine a preset result weight information group meeting a preset weight information order condition in a preset result weight information group sequence as the target result weight information group. The preset weight information sequence condition may be that the sequence number of the preset result weight information group in the preset result weight information group sequence is the smallest. The target result weight information set may represent a preset result weight information set with a minimum sequence number in a preset result weight information set sequence. The target result weight information set may represent the weight of the result output by the target user early warning information generation model and the weight of the result output by the user early warning information generation model to be switched. For example, the target result weight information set may characterize that the weight of the result output by the target user early warning information generation model is 20% and the weight of the result output by the user early warning information generation model to be switched is 80%.

Step 105, according to the preset result weight information group sequence and the target result weight information group, for the user early warning information generation model to be switched, executing the following switching steps:

step 1051, deleting the target result weight information group from the preset result weight information group sequence to update the preset result weight information group sequence.

In some embodiments, the executing entity may delete the target result weight information set from the preset result weight information set sequence to update the preset result weight information set sequence. Thus, the preset result weight information group sequence can be updated.

In step 1052, in response to detecting the input request information of the service data information, and the target result weight information set satisfies the preset result weight condition, the service data information is input to the to-be-switched user early warning information generation model and the target user early warning information generation model, so as to obtain the output result information of the to-be-switched user early warning information generation model as the first output result information and the output result information of the target user early warning information generation model as the second output result information.

In some embodiments, in response to detecting input request information for service data information, where the target result weight information set satisfies a preset result weight condition, the executing body may input the service data information into the to-be-switched user early-warning information generation model and the target user early-warning information generation model, and obtain output result information of the to-be-switched user early-warning information generation model as first output result information and output result information of the target user early-warning information generation model as second output result information. The preset result weight condition may be that the weight of the result output by the target user early warning information generation model represented by the target result weight information set and the weight of the result output by the user early warning information generation model to be switched are both not 0.

And 1053, according to the target result weight information group, combining the first output result information and the second output result information to obtain service data output result information corresponding to the service data information, so as to perform switching processing on the early warning information generation model of the user to be switched.

In some embodiments, the executing body may perform a combination process on the first output result information and the second output result information according to the target result weight information set to obtain service data output result information corresponding to the service data information, so as to perform a switching process on the to-be-switched user early warning information generation model. The service data output result information may be an output result of the corresponding service data information. In practice, the execution main body can perform combination processing on the first output result information and the second output result information according to the target result weight information set in various modes to obtain service data output result information corresponding to the service data information so as to perform switching processing on the early warning information generation model of the user to be switched.

In some optional implementations of some embodiments, the executing body may perform a combination process on the first output result information and the second output result information according to the target result weight information set to obtain service data output result information corresponding to the service data information, so as to perform a switching process on the to-be-switched user early warning information generation model by:

And a first step of determining target result weight information corresponding to the first output result information from the target result weight information group as first target result weight information. The first target result weight information may represent a weight of a result of the output of the user early warning information generation model to be switched. In practice, the execution subject may determine, as the first target result weight information, a weight representing a result of the output of the user early warning information generation model to be switched in the target result weight information group.

And a second step of determining target result weight information corresponding to the second output result information from the target result weight information group as second target result weight information. The second target result weight information may represent a weight of a result of the output of the user early warning information generation model to be switched. In practice, the execution subject may determine, as the second target result weight information, a weight representing a result of the output of the target user early-warning information generation model in the target result weight information group.

And thirdly, generating service data output result information corresponding to the service data information according to the determined first target result weight information and the determined second target result weight information. In practice, the executing body may multiply the weight represented by the first target result weight information with the first output result information corresponding to the service data information to obtain a first product. And then multiplying the determined weight represented by the second target result weight information with the second output result information corresponding to the service data information to obtain a second product. And finally, determining the average value of the obtained first product and the second product as service data output result information of the corresponding service data information.

Optionally, after step 1053, the executing body may further execute the switching step again in response to detecting that the change interval time information corresponding to the target result weight information group satisfies the preset time interval condition, and the updated preset result weight information group sequence is not null, taking the preset result weight information group satisfying the preset weight information order condition in the updated preset result weight information group sequence as the target result weight information group, and taking the updated preset result weight information group sequence as the preset result weight information group sequence. The change interval time information may represent a time interval between last updating of the target result weight information. For example, the change interval time information may be 10 minutes. The preset time interval condition may be that the time interval represented by the change interval time information is the same as the preset time interval. Here, the specific value of the preset time interval is not particularly limited.

Optionally, after step 1053, the executing body may further input the service data information to the target user early warning information generation model in response to detecting that the input request information for the service data information does not meet the preset result weight condition, so as to obtain output result information of the target user early warning information generation model as service data output result information corresponding to the service data information, so as to complete the switching process of the user early warning information generation model to be switched. The input request information may be a request for inputting service data information.

Alternatively, after the step 1053, first, the executing entity may determine whether the service data output result information meets a preset early warning condition according to the service data output result information. The default pre-warning condition may be that the default rate represented by the service data output result information is greater than a preset default rate value. Here, the specific value of the preset default rate value is not particularly limited.

And then, in response to determining that the service data output result information meets the preset early warning condition, determining early warning grade information corresponding to the service data information according to a preset early warning grade information set. The early warning level information may represent an early warning level corresponding to the service data output result information. The preset early warning level information in the preset early warning level information set can represent the corresponding relation between the service data output result information and the early warning level. For example, the alert level information may be: "service data output result information: 70%, early warning grade information: and (5) secondary early warning. The early warning grade information can represent the early warning grade of the service data output result information. In practice, in response to determining that the service data output result information meets the preset early warning condition, the executing body may determine early warning level information included in preset early warning level information, where the preset early warning level information is the same as service data output result information corresponding to the service data information, included in the preset early warning level information set, as early warning level information corresponding to the service data information.

And then, the preset early warning level mode information corresponding to the early warning level information in the preset early warning level mode information set can be determined to be target preset early warning level mode information. The preset pre-warning level mode information in the preset pre-warning level mode information set can represent a corresponding relation between the pre-warning level information and the pre-warning mode. For example, the preset early warning level mode information may be: "early warning level information: secondary early warning; early warning mode: short message early warning. In practice, the executing body may determine the preset early warning level mode information, which is included in the preset early warning level mode information set and is the same as the early warning level mode information, as the target preset early warning level mode information.

Finally, the business data information can be subjected to early warning processing according to the target preset early warning level mode information. In practice, the executing body may perform early warning processing on the service data information according to an early warning mode included in the target preset early warning level mode information. For example, the pre-warning mode included in the target pre-set pre-warning level mode information may be a short message pre-warning, and the executing body may send the pre-set short message template information to a terminal device for pre-warning processing. For example, the preset sms template information may be: "service data information: xx is at a higher risk and please improve vigilance.

With further reference to fig. 2, as an implementation of the method shown in the foregoing figures, the present disclosure provides some embodiments of a service model switching method, where the apparatus embodiments correspond to those method embodiments shown in fig. 1, and the apparatus may be specifically applied to various electronic devices.

As shown in fig. 2, the switching apparatus 200 of the service model of some embodiments includes: a detection unit 201, a first determination unit 202, a second determination unit 203, a third determination unit 204, and a switching unit 205. The detection unit 201 is configured to perform performance detection processing on a to-be-switched user early-warning information generation model and a target user early-warning information generation model to obtain service model performance information corresponding to the to-be-switched user early-warning information generation model and service model performance information corresponding to the target user early-warning information generation model, where the target user early-warning information generation model is used for replacing the to-be-switched user early-warning information generation model; a first determining unit 202 configured to determine, according to the service model performance information of the user early warning information generation model to be switched, whether the user early warning information generation model to be switched meets a preset fast switching condition; a second determining unit 203 configured to determine whether preset switching requirement information corresponding to the user early warning information generation model to be switched meets a preset fast switching requirement condition in response to determining that the user early warning information generation model to be switched does not meet the preset fast switching condition; a third determining unit 204 configured to determine, as a target result weight information group, a preset result weight information group satisfying a preset weight information order condition in a preset result weight information group sequence in response to determining that preset switching requirement information corresponding to the user early warning information generation model to be switched does not satisfy the preset fast switching requirement condition; a switching unit 205 configured to perform the following switching steps for the user early warning information generation model to be switched according to the preset result weight information group sequence and the target result weight information group: deleting the target result weight information group from the preset result weight information group sequence to update the preset result weight information group sequence; in response to detecting input request information of service data information, and the target result weight information set meets a preset result weight condition, inputting the service data information into the to-be-switched user early warning information generation model and the target user early warning information generation model to obtain output result information of the to-be-switched user early warning information generation model as first output result information and output result information of the target user early warning information generation model as second output result information; and according to the target result weight information group, carrying out combination processing on the first output result information and the second output result information to obtain service data output result information corresponding to the service data information, so as to carry out switching processing on the early warning information generation model of the user to be switched.

It will be appreciated that the elements described in the apparatus 200 correspond to the various steps in the method described with reference to fig. 1. Thus, the operations, features and resulting benefits described above for the method are equally applicable to the apparatus 200 and the units contained therein, and are not described in detail herein.

Referring now to fig. 3, a schematic diagram of an electronic device 300 (e.g., a computing device) suitable for use in implementing some embodiments of the present disclosure is shown. The electronic device shown in fig. 3 is merely an example and should not impose any limitations on the functionality and scope of use of embodiments of the present disclosure.

As shown in fig. 3, the electronic device 300 may include a processing means 301 (e.g., a central processing unit, a graphics processor, etc.) that may perform various suitable actions and processes in accordance with a program stored in a Read Only Memory (ROM) 302 or a program loaded from a storage means 308 into a Random Access Memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic apparatus 300 are also stored. The processing device 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input/output (I/O) interface 305 is also connected to bus 304.

In general, the following devices may be connected to the I/O interface 305: input devices 306 including, for example, a touch screen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; an output device 307 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, and the like; storage 308 including, for example, magnetic tape, hard disk, etc.; and communication means 309. The communication means 309 may allow the electronic device 300 to communicate with other devices wirelessly or by wire to exchange data. While fig. 3 shows an electronic device 300 having various means, it is to be understood that not all of the illustrated means are required to be implemented or provided. More or fewer devices may be implemented or provided instead. Each block shown in fig. 3 may represent one device or a plurality of devices as needed.

In particular, according to some embodiments of the present disclosure, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, some embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method shown in the flow chart. In such embodiments, the computer program may be downloaded and installed from a network via communications device 309, or from storage device 308, or from ROM 302. The above-described functions defined in the methods of some embodiments of the present disclosure are performed when the computer program is executed by the processing means 301.

It should be noted that, the computer readable medium described in some embodiments of the present disclosure may be a computer readable signal medium or a computer readable storage medium, or any combination of the two. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, 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. In some embodiments of the present disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In some embodiments of the present disclosure, however, the computer-readable signal medium may comprise a data signal propagated in baseband or as part of a carrier wave, with the computer-readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, fiber optic cables, RF (radio frequency), and the like, or any suitable combination of the foregoing.

In some embodiments, the clients, servers may communicate using any currently known or future developed network protocol, such as HTTP (HyperText Transfer Protocol ), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), the internet (e.g., the internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future developed networks.

The computer readable medium may be contained in the electronic device; or may exist alone without being incorporated into the electronic device. The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to: performing performance detection processing on a user early warning information generation model to be switched and a target user early warning information generation model to obtain service model performance information corresponding to the user early warning information generation model to be switched and service model performance information corresponding to the target user early warning information generation model, wherein the target user early warning information generation model is used for replacing the user early warning information generation model to be switched; determining whether the early warning information generation model of the user to be switched meets a preset quick switching condition according to the service model performance information of the early warning information generation model of the user to be switched; in response to determining that the user early warning information generation model to be switched does not meet the preset quick switching condition, determining whether preset switching requirement information corresponding to the user early warning information generation model to be switched meets the preset quick switching requirement condition; in response to determining that the preset switching requirement information corresponding to the user early warning information generation model to be switched does not meet the preset quick switching requirement condition, determining a preset result weight information group meeting the preset weight information sequence condition in a preset result weight information group sequence as a target result weight information group; according to the preset result weight information group sequence and the target result weight information group, for the user early warning information generation model to be switched, executing the following switching steps: deleting the target result weight information group from the preset result weight information group sequence to update the preset result weight information group sequence; in response to detecting input request information of service data information, and the target result weight information set meets a preset result weight condition, inputting the service data information into the to-be-switched user early warning information generation model and the target user early warning information generation model to obtain output result information of the to-be-switched user early warning information generation model as first output result information and output result information of the target user early warning information generation model as second output result information; and according to the target result weight information group, carrying out combination processing on the first output result information and the second output result information to obtain service data output result information corresponding to the service data information, so as to carry out switching processing on the early warning information generation model of the user to be switched.

Computer program code for carrying out operations for some embodiments of the present disclosure may be written in one or more programming languages, including an object oriented programming language such as Java, smalltalk, C ++ and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (for example, through the Internet using an Internet service provider).

The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

The units described in some embodiments of the present disclosure may be implemented by means of software, or may be implemented by means of hardware. The described units may also be provided in a processor, for example, described as: the device comprises a detection unit, a first determination unit, a second determination unit, a third determination unit and a switching unit. The names of the units do not form a limitation on the unit itself in a certain case, for example, the detection unit may also be described as a "unit for performing performance detection processing on the to-be-switched user early-warning information generation model and the target user early-warning information generation model to obtain service model performance information corresponding to the to-be-switched user early-warning information generation model and service model performance information corresponding to the target user early-warning information generation model".

The functions described above herein may be performed, at least in part, by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: a Field Programmable Gate Array (FPGA), an Application Specific Integrated Circuit (ASIC), an Application Specific Standard Product (ASSP), a system on a chip (SOC), a Complex Programmable Logic Device (CPLD), and the like.

The foregoing description is only of the preferred embodiments of the present disclosure and description of the principles of the technology being employed. It will be appreciated by those skilled in the art that the scope of the invention in the embodiments of the present disclosure is not limited to the specific combination of the above technical features, but encompasses other technical features formed by any combination of the above technical features or their equivalents without departing from the spirit of the invention. Such as the above-described features, are mutually substituted with (but not limited to) the features having similar functions disclosed in the embodiments of the present disclosure.

Claims

1. A switching method of service models comprises the following steps:

Performing performance detection processing on a user early warning information generation model to be switched and a target user early warning information generation model to obtain service model performance information corresponding to the user early warning information generation model to be switched and service model performance information corresponding to the target user early warning information generation model, wherein the target user early warning information generation model is used for replacing the user early warning information generation model to be switched;

Determining whether the early warning information generation model of the user to be switched meets a preset quick switching condition according to the service model performance information of the early warning information generation model of the user to be switched;

Determining whether preset switching requirement information corresponding to the user early warning information generation model to be switched meets preset quick switching requirement conditions or not in response to determining that the user early warning information generation model to be switched does not meet the preset quick switching conditions;

In response to determining that the preset switching requirement information corresponding to the user early warning information generation model to be switched does not meet the preset quick switching requirement condition, determining a preset result weight information group meeting a preset weight information sequence condition in a preset result weight information group sequence as a target result weight information group;

according to a preset result weight information group sequence and a target result weight information group, for the user early warning information generation model to be switched, executing the following switching steps:

deleting the target result weight information group from the preset result weight information group sequence to update the preset result weight information group sequence;

Responding to the detection of input request information of service data information, and inputting the service data information into the to-be-switched user early warning information generation model and the target user early warning information generation model, wherein a target result weight information set meets a preset result weight condition, so that output result information of the to-be-switched user early warning information generation model is obtained as first output result information, and output result information of the target user early warning information generation model is obtained as second output result information;

And according to the target result weight information group, carrying out combination processing on the first output result information and the second output result information to obtain service data output result information corresponding to the service data information, so as to carry out switching processing on the early warning information generation model of the user to be switched.

2. The method of claim 1, wherein the step of switching further comprises:

And in response to detecting that the change interval time information of the corresponding target result weight information group meets a preset time interval condition, and the updated preset result weight information group sequence is not null, taking the updated preset result weight information group meeting the preset weight information sequence condition in the updated preset result weight information group sequence as the target result weight information group, and taking the updated preset result weight information group sequence as the preset result weight information group sequence, and executing the switching step again.

3. The method of claim 1, wherein the step of switching further comprises:

And in response to detecting that the input request information of the service data information is detected, and the target result weight information group does not meet the preset result weight condition, inputting the service data information into the target user early warning information generation model, and obtaining output result information of the target user early warning information generation model as service data output result information corresponding to the service data information so as to complete switching processing of the user early warning information generation model to be switched.

4. The method of claim 1, wherein performing performance detection on the to-be-switched user early-warning information generation model and the target user early-warning information generation model to obtain service model performance information corresponding to the to-be-switched user early-warning information generation model and service model performance information corresponding to the target user early-warning information generation model, comprises:

acquiring a historical service data information set;

Performing data cleaning processing on each history service data information in the acquired history service data information set to obtain a history service data information set after the data cleaning processing as a target history service data information set;

Inputting each target historical service data information in the target historical service data information set into the to-be-switched user early warning information generation model and the target user early warning information generation model respectively;

determining each piece of prediction result information corresponding to the target historical service data information set output by the user early warning information generation model to be switched as each piece of first prediction result information, wherein the first prediction result information in each piece of first prediction result information corresponds to the target historical service data information in the target historical service data information set;

Determining each piece of prediction result information corresponding to the target historical service data information set output by the target user early warning information generation model as each piece of second prediction result information, wherein the second prediction result information in each piece of second prediction result information corresponds to the target historical service data information in the target historical service data information set;

And generating service model performance information corresponding to the user early warning information generation model to be switched and service model performance information corresponding to the target user early warning information generation model according to the actual early warning result information set corresponding to the target historical service data information set, the first prediction result information and the second prediction result information.

5. The method of claim 4, wherein the generating business model performance information corresponding to the user early warning information generation model to be switched and business model performance information corresponding to the target user early warning information generation model according to the actual early warning result information set corresponding to the target historical business data information set, the respective first prediction result information, and the respective second prediction result information comprises:

determining each piece of first prediction result information output by the user early warning information generation model to be switched as each piece of target prediction result information;

For each of the respective target prediction result information, the following detection steps are performed:

determining target historical service data information corresponding to the target prediction result information in the target historical service data information set as historical service data information;

determining actual early warning result information corresponding to the historical service data information in the actual early warning result information set as target actual early warning result information;

Performing similarity detection processing on the determined target actual early warning result information and the target prediction result information to obtain similarity information corresponding to the determined target actual early warning result information and the target prediction result information;

Taking each piece of second predicted result information output by the target user early-warning information generation model as each piece of target predicted result information, and executing the detection step again for each piece of target predicted result information in each piece of target predicted result information to obtain each piece of similarity information corresponding to the target user early-warning information generation model;

and generating service model performance information corresponding to the user early warning information generation model to be switched and service model performance information corresponding to the target user early warning information generation model according to the similarity information corresponding to the user early warning information generation model to be switched, the similarity information corresponding to the target user early warning information generation model and the actual early warning result information set.

6. The method of claim 1, wherein the combining the first output result information and the second output result information according to the target result weight information set to obtain service data output result information corresponding to service data information, so as to perform switching processing on the to-be-switched user early warning information generation model, includes:

Determining target result weight information corresponding to the first output result information from a target result weight information group as first target result weight information;

determining target result weight information corresponding to the second output result information from the target result weight information group as second target result weight information;

And generating service data output result information corresponding to the service data information according to the determined first target result weight information and the determined second target result weight information.

7. The method of claim 6, wherein the method further comprises:

Determining whether the service data output result information meets a preset early warning condition according to the service data output result information;

responding to the fact that the service data output result information meets the preset early warning condition, and determining early warning grade information corresponding to the service data information according to a preset early warning grade information set;

Determining preset early warning grade mode information corresponding to the early warning grade information in a preset early warning grade mode information set as target preset early warning grade mode information;

And carrying out early warning processing on the business data information according to the target preset early warning level mode information.

8. A switching device for a service model, comprising:

The detection unit is configured to perform performance detection processing on a user early warning information generation model to be switched and a target user early warning information generation model to obtain service model performance information corresponding to the user early warning information generation model to be switched and service model performance information corresponding to the target user early warning information generation model, wherein the target user early warning information generation model is used for replacing the user early warning information generation model to be switched;

The first determining unit is configured to determine whether the early warning information generation model of the user to be switched meets a preset quick switching condition according to the service model performance information of the early warning information generation model of the user to be switched;

A second determining unit configured to determine whether preset switching requirement information corresponding to the user early warning information generation model to be switched meets a preset quick switching requirement condition in response to determining that the user early warning information generation model to be switched does not meet the preset quick switching condition;

A third determining unit configured to determine, as a target result weight information group, a preset result weight information group satisfying a preset weight information order condition in a preset result weight information group sequence in response to determining that preset switching requirement information corresponding to the user early warning information generation model to be switched does not satisfy the preset fast switching requirement condition;

The switching unit is configured to execute the following switching steps for the user early warning information generation model to be switched according to a preset result weight information group sequence and a target result weight information group: deleting the target result weight information group from the preset result weight information group sequence to update the preset result weight information group sequence; responding to the detection of input request information of service data information, and inputting the service data information into the to-be-switched user early warning information generation model and the target user early warning information generation model, wherein a target result weight information set meets a preset result weight condition, so that output result information of the to-be-switched user early warning information generation model is obtained as first output result information, and output result information of the target user early warning information generation model is obtained as second output result information; and according to the target result weight information group, carrying out combination processing on the first output result information and the second output result information to obtain service data output result information corresponding to the service data information, so as to carry out switching processing on the early warning information generation model of the user to be switched.

9. An electronic device, comprising:

one or more processors;

a storage device having one or more programs stored thereon;

when executed by the one or more processors, causes the one or more processors to implement the method of any of claims 1 to 7.

10. A computer readable medium having stored thereon a computer program, wherein the computer program, when executed by a processor, implements the method of any of claims 1 to 7.