Decision model-based activity decision generation method, apparatus and device, and medium

Abstract

The invention discloses an activity decision generation method and device based on a decision model, equipment and a medium. The method is applied to an intelligent decision-making model, and comprises the following steps: acquiring corresponding actual parameters and standard parameters from a preset data source according to service requirements; matching a corresponding target deviation algorithm according to the actual parameter and the standard parameter; calculating a deviation value between the actual parameter and the standard parameter according to the target deviation algorithm; and matching a corresponding event value according to the deviation value, and determining a corresponding execution activity according to the event value. By implementing the method provided by the invention, the problem that decision judgment cannot be efficiently and automatically performed in the prior art can be solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent decision-making, and in particular to an activity decision generation method and device based on a decision model, equipment and a medium. BACKGROUND

[0002] At present, the decision-making system of many organizations still takes human managers as the center, relies on experience accumulation and structured analysis framework, and highly depends on the subjective judgment, value orientation and comprehensive judgment ability of managers on complex environment. The system presents obvious hierarchical division of labor and process characteristics, strictly follows the preset rules and steps, and needs multi-department and multi-level coordination. However, this traditional mode relies on artificial transmission of information at each level, which leads to information distortion and inefficient process, and the decision-making response speed lags behind, which makes it difficult to adapt to the needs of organizational digital transformation and real-time online operation, and to cope with complex business scenarios, so that the prior art cannot efficiently and automatically make decision-making. SUMMARY

[0003] The embodiments of the present application provide an activity decision generation method and device based on a decision model, equipment and a medium, aiming at solving the problem that the prior art cannot efficiently and automatically make decision-making.

[0004] In a first aspect, the embodiments of the present application provide an activity decision generation method based on a decision model, applied to an intelligent decision-making model, which includes: acquiring corresponding actual parameters and standard parameters in a preset data source according to business needs; matching a corresponding target deviation algorithm according to the actual parameters and the standard parameters; calculating a deviation value between the actual parameters and the standard parameters according to the target deviation algorithm; matching a corresponding event value according to the deviation value, and determining a corresponding execution activity according to the event value.

[0005] In a second aspect, the embodiments of the present application also provide an activity decision generation device based on a decision model, which is applied to an intelligent decision-making model, and includes: an acquisition unit, configured to acquire corresponding actual parameters and standard parameters in a preset data source according to business needs; a matching unit, configured to match a corresponding target deviation algorithm according to the actual parameters and the standard parameters; a calculation unit, configured to calculate a deviation value between the actual parameters and the standard parameters according to the target deviation algorithm; and a determination unit, configured to match a corresponding event value according to the deviation value, and determine a corresponding execution activity according to the event value.

[0006] In a third aspect, the embodiments of the present application also provide a computer device, which includes a memory and a processor, and the memory stores a computer program, and the processor implements the above method when executing the computer program.

[0007] In a fourth aspect, the embodiments of the present application further provide a computer readable storage medium, which stores a computer program. The computer program includes program instructions, and the program instructions can implement the above method when executed by a processor.

[0008] The embodiments of the present application provide a decision model-based activity decision generation method and device, equipment and medium, which are applied to intelligent decision model, and the method comprises the following steps: acquiring corresponding actual parameters and standard parameters from a preset data source according to business requirements; matching a corresponding target deviation algorithm according to the actual parameters and the standard parameters; calculating a deviation value between the actual parameters and the standard parameters according to the target deviation algorithm; matching a corresponding event value according to the deviation value, and determining a corresponding execution activity according to the event value. The embodiments of the present application acquire corresponding actual parameters and standard parameters from a preset data source, and automatically match a corresponding deviation algorithm to acquire a highly adaptive deviation algorithm, thereby avoiding result distortion caused by algorithm mismatch. The business activities to be executed are determined from a pre-defined activity library according to the calculated deviation value, so as to realize automatic decision, improve decision efficiency and accuracy, and adapt to complex and changeable business scenarios. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0010] Figure 1 The flowchart of the decision model-based activity decision generation method provided by the embodiments of the present application is shown in the figure. Figure 2 The first sub-flowchart of the decision model-based activity decision generation method provided by the embodiments of the present application is shown in the figure. Figure 3 The second sub-flowchart of the decision model-based activity decision generation method provided by the embodiments of the present application is shown in the figure. Figure 4 The third sub-flowchart of the decision model-based activity decision generation method provided by the embodiments of the present application is shown in the figure. Figure 5 The fourth sub-flowchart of the decision model-based activity decision generation method provided by the embodiments of the present application is shown in the figure. Figure 6 The fifth sub-flowchart of the decision model-based activity decision generation method provided by the embodiments of the present application is shown in the figure. Figure 7 The sixth sub-flowchart of the decision model-based activity decision generation method provided by the embodiments of the present application is shown in the figure. Figure 8 A schematic block diagram of an activity decision generation device based on a decision model is provided for an embodiment of the present application. Figure 9 A schematic block diagram of a computer device is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0011] The technical solutions in the embodiments of the present application will be clearly and completely described with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0012] It should be understood that, when used in the specification and the appended claims, the terms “comprise” and “include” indicate the presence of described features, integers, steps, operations, elements, and / or components, but do not exclude one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0013] It should also be understood that the terms used in the present application specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the present application specification and the appended claims, the singular forms “a”, “an” and “the” are intended to include the plural forms, unless the context clearly indicates otherwise.

[0014] It should be further understood that the term “and / or” used in the present application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0015] Please refer to Figure 1 , Figure 1 A flowchart of a decision model-based activity decision generation method is provided for an embodiment of the present application. The decision model-based activity decision generation method in the embodiment can be applied to an intelligent decision model, which can be deployed in an intelligent decision system or a corresponding APP. The intelligent decision model includes a perception module, an identification module, and a reaction module. The perception module acquires corresponding parameters, the identification module makes a decision judgment, and the reaction module matches a corresponding activity according to the decision judgment result of the identification module. An application party can subscribe to the intelligent decision model from a decision model market to provide decision services, for example, a hospital can subscribe to a decision model to provide corresponding diagnosis services according to examination data of a patient. By using the intelligent decision model with the method, automatic decision can be realized, the decision efficiency and accuracy are improved, and complex and variable business scenarios are adapted.

[0016] Figure 1 is a flowchart of the activity decision generation method based on the decision model provided by the embodiments of the present application. As shown in the figure, the method comprises the following steps S110-S140.

[0017] S110, obtaining corresponding actual parameters and standard parameters in a preset data source according to a business requirement.

[0018] In this embodiment, the business requirement is determined according to the requirement of the application party (user), for example, in the medical diagnosis field, the business requirement can be to judge whether the patient has hypertension according to the blood pressure data of the patient. The preset data source is a place capable of storing data. In this embodiment, the preset data source is a meta database, a knowledge database, an information database, an event knowledge database and a wisdom database provided by a data market. The actual parameter is the real data generated in the running process of the business, which can be obtained by subscribing from the preset data source. For example, in the medical diagnosis field, the actual parameter can be the systolic and diastolic pressure data of the patient, which can be measured by a medical device and uploaded to the database for the decision model to subscribe. The standard parameter is a benchmark value preset for measuring whether the actual parameter meets the requirement, which can also be obtained by subscribing from the preset data source, for example, in the medical diagnosis field, the standard parameter can be the diagnostic criteria of hypertension (such as systolic pressure greater than 140 mmHg or diastolic pressure greater than 90 mmHg), which can be formulated by medical experts and uploaded to the data market. The perception module of the intelligent decision model obtains corresponding actual parameters and standard parameters in the preset data source according to the business requirement, specifically, the analysis of the business requirement determines the required parameters, and the corresponding actual parameters and standard parameters are found or subscribed from the preset data source. By obtaining the corresponding actual parameters and standard parameters according to the business requirement, a data basis is provided for subsequent activity generation.

[0019] In an embodiment, as shown in Figure 2 , the step S110 further comprises steps S1101-S1103 before the step S110.

[0020] S1101, generating a corresponding training execution activity according to the training actual parameters and the training standard parameters in the training data set; S1102, comparing the training execution activity with the training label activity in the training data set to generate error information; S1103, adjusting the corresponding model parameter value according to the error information to determine the intelligent decision model.

[0021] In the embodiment, after the developer builds the decision model instance, the developer needs to train the decision model instance and then put it into use. Specifically, the developer configures a target value, an actual value, a deviation algorithm, an event (value), and an execution activity. The target value, the actual value, and the event value are subscribed in a preset data source, and the deviation algorithm is subscribed from an algorithm market. After each module completes the subscription and the configuration, the editing of a decision model instance is completed. When editing, the basic information of the decision model needs to be filled in, such as a name, an industry to which the decision model belongs, a function description, and a note information. Then, the training data is input into the edited decision model, and the decision model is trained according to the training data. Specifically, according to the training actual parameter and the training standard parameter in the training data set, the corresponding training execution activity is generated. The decision model calculates a deviation value according to the training actual parameter and the training standard parameter, and matches the corresponding event value according to the deviation value to determine the corresponding training execution activity. The training execution activity is compared with the training label activity corresponding to the training actual parameter and the training standard parameter in the training data set to obtain error information. According to the error information, the corresponding model parameter value is adjusted. Specifically, the model parameter value can be adjusted and optimized according to the error information by using a loss function, an error function, and the like, so as to minimize the error information. For example, the error information is propagated from an output layer to an input layer by using back propagation, the gradient of each parameter is calculated, and the parameter value is updated, so as to obtain a model with the best performance, which is determined as the intelligent decision model. The model is learned by using the training data, and the parameters of the model are gradually optimized, so as to improve the accuracy of decision of the model. In the embodiment, when the intelligent decision model meets the training requirement, the intelligent decision model enters a publishing state. Before publishing, whether each configuration is legal needs to be detected, including whether the continuity of a deviation item and the completeness of an algorithm input parameter are configured. A version number is generated, and after the detection is legal, the intelligent decision model can be published to a decision model market and is visible to an application layer. The application layer can view and subscribe the intelligent decision model by using the decision model market. When the application layer calls the intelligent decision model, the decision model software starts to execute S110, and the subsequent steps are executed.

[0022] In an embodiment, as shown in FIG. 11, the step S1103 further includes steps S11031-S11032. Figure 3

[0023] S11031, determining whether the accuracy of the training execution activity generated by the adjusted intelligent decision model reaches a preset accuracy; S11032, if not, continuing to adjust the model parameter value according to the training data set.

[0024] ​In the embodiment, the accuracy is an important indicator for measuring the performance of the model, representing the proportion of the number of samples predicted correctly by the model on the training dataset to the total number of samples. The preset accuracy can be set according to business requirements or application requirements, and is not limited. After adjusting the model parameters each time, the adjusted model is used to predict the training dataset, to generate training execution activities, which are compared with the training label activities to determine whether they are correct or not, so as to calculate the accuracy of the training execution activities. The calculated accuracy is compared with the preset normal rate. If the calculated accuracy is lower than the preset normal rate, it means that the performance of the model under the current parameters has not reached the expectation, and the model parameter values are continued to be adjusted according to the training dataset, for example, the gradient of the parameter is calculated according to the loss function (such as mean square error, cross entropy, etc.), and the parameter value is updated accordingly. After adjusting the parameters, the model is used again to predict the training dataset, and the new accuracy is calculated, until the accuracy of the model reaches or exceeds the preset normal rate. By judging whether the accuracy of the training execution activities generated by the adjusted intelligent decision model reaches the preset normal rate, and deciding whether to continue adjusting the model parameter values according to the comparison result, it can be ensured that the model is continuously optimized in the training process, and finally reaches the satisfactory performance standard.

[0025] S120, according to the actual parameter and the standard parameter matching corresponding target deviation algorithm.

[0026] In the embodiment, the target deviation algorithm is an algorithm for calculating the deviation between the actual parameter and the standard parameter. It can be understood that different business scenarios and data types may require different deviation algorithms for calculation. For example, for numerical data, a simple subtraction algorithm can be used to calculate the deviation; and for more complex data types or business scenarios, advanced algorithms such as neural networks, large models, etc. may be required for deviation calculation. Therefore, the target deviation algorithm corresponding to the actual parameter and the standard parameter needs to be matched, specifically, the perception module can automatically select or match the most suitable deviation algorithm from the algorithm market according to the attribute information (such as data type, range, precision, etc.) of the actual parameter and the standard parameter and the label and input requirements of the deviation algorithm. For example, if the actual parameter and the standard parameter are the numerical values corresponding to the blood pressure of a patient, since the blood pressure data is numerical data, and the size relationship between the actual value and the standard value needs to be compared, a simple subtraction algorithm can be selected as the target deviation algorithm. By automatically matching the most suitable target deviation algorithm according to the characteristics of the actual parameter and the standard parameter, the accuracy and effectiveness of the deviation calculation are ensured.

[0027] In an embodiment, as shown in FIG. 12, the step S120 further includes steps S121-S122. Figure 4

[0028] ​S121, filtering out an initial deviation algorithm set within a preset similarity range according to attribute information of the actual parameter and the standard parameter; S122, filtering out the target deviation algorithm from the initial deviation algorithm set according to data types of the actual parameter and the standard parameter.

[0029] In this embodiment, the perception module filters the target deviation algorithm in the algorithm market. It should be noted that each deviation algorithm is labeled with attribute tags related to it. These tags describe the data types and business scenarios to which the algorithm is applicable. Therefore, the actual parameter and the standard parameter can be filtered according to the attribute information in the algorithm market. For example, the perception module analyzes the attribute information of the actual blood pressure data and the diagnostic standard, such as digital data and medical diagnosis business field. Calculate the similarity of these attribute information and the labels of each deviation algorithm in the algorithm market. For example, algorithms related to medical diagnosis and numerical data processing algorithms have a higher similarity. Filter out algorithms within a preset range of similarity, such as subtraction algorithm, mean algorithm, neural network algorithm, etc., to form an initial deviation algorithm set. The preset similarity range can be set according to specific application requirements, which is not limited. According to the data type of the actual parameter and the standard parameter, the target deviation algorithm is filtered out from the initial deviation algorithm set, wherein the data type is the specific form of the data, such as numerical type, boolean type, string type, etc. Filter the deviation algorithm that best fits the data type in the initial deviation algorithm set. For example, if the number is a numerical type, the subtraction algorithm is used as the target deviation algorithm. By filtering the most suitable target deviation algorithm according to the attribute information and data type of the actual parameter and the standard parameter, a highly adaptive deviation algorithm is obtained, avoiding the distortion of the result caused by the mismatch of the algorithm.

[0030] S130, calculating the deviation value between the actual parameter and the standard parameter according to the target deviation algorithm.

[0031] In this embodiment, the deviation value calculation is to input the actual parameter into the target deviation algorithm and compare it with the standard parameter to obtain the difference between them. According to the target deviation algorithm, the deviation value between the actual parameter and the standard parameter is calculated. Specifically, the recognition module inputs the actual parameter and the standard parameter into the target deviation algorithm, and the target deviation algorithm processes the input data according to its internal logic (such as subtraction, mean calculation, neural network inference, etc.), calculates and outputs the deviation value, wherein the output deviation value has a corresponding positive or negative value, for example, the input data is the actual systolic pressure: 130mmHg, the standard systolic pressure: 140mmHg, and the output deviation value is - (negative) 10mmHg. By calculating the deviation value between the actual parameter and the standard parameter, accurate reference data is provided for subsequent decision analysis.

[0032] S140, according to the deviation value matches the corresponding event value, according to the event value determines the corresponding execution activity.

[0033] In this embodiment, the event value is a classification or labeled representation of the event, which is used to identify the events corresponding to different business scenarios or states. Specifically, each event is provided with an event value, and each event value is bound to a preset deviation interval. According to the deviation value matches the corresponding event value, specifically, the recognition module according to the deviation value and the corresponding event value match, for example, the deviation value of systolic pressure is between 0-15mmHg and the deviation value of diastolic pressure is between 0-10mmHg, then the event value determined according to the deviation range is 1, which is "mild hypertension". Then it is fed back to the reaction module. The reaction module determines the corresponding execution activity according to the event value, and the reaction module matches the specific execution activity according to the event value output by the recognition module, and the execution activity is a specific operation or response measure defined according to the event value. For example, for the "mild hypertension" event value, the execution activity may be to suggest lifestyle adjustment. It should be noted that the developer pre-configures the mapping relationship between the event (value) and the execution activity in the decision model, and the decision model finally returns the activity corresponding to the event after the event is triggered. From now on, the decision of the business activity is completed. By matching the corresponding execution activity according to the matched event value, automatic decision is realized, and the decision efficiency and accuracy are improved.

[0034] In an embodiment, as shown in Figure 5 The step S140 includes steps S141-S142.

[0035] S141, according to the deviation value in the preset deviation interval, obtain the target deviation interval; S142, according to the target deviation interval determines the corresponding event value.

[0036] In the embodiment, the preset deviation interval is a numerical range defined in advance according to business requirements and domain knowledge, and is used for classifying the deviation value. For example, in the diagnosis of hypertension, different blood pressure deviation intervals can be set, such as "normal range", "mild hypertension", "moderate hypertension", "severe hypertension", etc. According to the matching of the deviation value in the preset deviation interval, the target deviation interval is obtained. Specifically, the recognition module compares the calculated deviation value with the preset deviation interval, determines which interval the deviation value falls into, and the deviation interval falling into is the target deviation interval, which identifies the category or severity to which the current deviation value belongs. According to the target deviation interval, the corresponding event value is determined. Specifically, the deviation interval is bound with the event in advance, and an event value is assigned to each event. Therefore, according to the target deviation interval, the corresponding event value can be determined. By matching the calculated deviation value with the preset deviation interval to determine the target deviation interval, the corresponding event value is determined, so as to ensure that different situations can be dynamically classified and responded according to the size and range of the deviation value.

[0037] In an embodiment, as shown in FIG. 14, the step S141 further includes steps S1411-S1413. Figure 6

[0038] S1411, determining the type of the deviation value, wherein the deviation value includes positive deviation, zero deviation, and negative deviation; S1412, screening the deviation value and its type in a plurality of preset first-level intervals to determine a target first-level interval; S1413, screening the deviation value in a plurality of preset second-level labels in the target first-level interval to determine the target deviation interval.

[0039] ​In this embodiment, the deviation algorithm outputs the positive and negative values of the deviation value when outputting the deviation value, so that the type of the deviation value can be directly determined, wherein the deviation value includes positive deviation (indicating that the actual parameter is higher than the standard parameter), zero deviation (indicating that the actual parameter is equal to the standard parameter), and negative deviation (indicating that the actual parameter is lower than the standard parameter). According to the deviation value and its type, the deviation value is screened in a plurality of preset first-level intervals to determine a target first-level interval. Specifically, the first-level interval is an interval defined in advance according to the type and approximate range of the deviation value, and is used for preliminary classification of the deviation value. The recognition module matches the deviation value with the preset first-level interval according to the type of the deviation value to determine the target first-level interval. For example, if the deviation value is positive, it is matched to a "positive deviation interval". The first-level interval includes a plurality of preset second-level labels to refine the classification of the deviation value. These second-level labels are usually based on business requirements and domain knowledge, and are used to identify different business scenarios or states. For example, within the "positive deviation interval", the following second-level labels can be set: mild positive deviation, moderate positive deviation, and severe positive deviation. The recognition module matches the deviation value with the preset second-level label according to the specific value of the deviation value within the target first-level interval to determine a target deviation interval. By determining the target deviation interval according to the deviation value, accurate decision-making can be achieved.

[0040] In an embodiment, as shown in FIG. 14B, the step S142 includes steps S1421-S1422. Figure 7

[0041] S1421, performing logical operation according to the deviation value to obtain an operation result; S1422, determining the event value corresponding to the target deviation interval according to the operation result.

[0042] In this embodiment, the logical operation is a combination analysis of a plurality of deviation items, for example, a complex conditional expression is constructed using logical operators (such as AND, OR). The deviation value can include deviation values of various data, so it is necessary to more accurately determine whether the combination of deviation items meets the condition of triggering an event. According to the deviation value, a logical operation is performed to obtain an operation result. Specifically, for example, the operation result is true only when all event values corresponding to the deviation values are the same. The specific logical operator used can be set according to the specific scene, which is not limited. According to the operation result, the event value corresponding to the target deviation interval is determined. Only when the operation result is true, the event value corresponding to the target deviation interval can be triggered, for example, when the deviation value A (blood pressure deviation > 10 mmHg) AND the deviation value B (heart rate deviation > 5 times / minute) both correspond to the same event value, the event value is determined as the final event value corresponding to the target deviation interval. By performing logical operation on the deviation value, the final event value is determined, so that accurate decision-making can still be achieved in complex scenarios.​

[0043] To further understand the activity decision generation method based on the decision model of the present application, the following is described by the processing flow of the activity decision: The intelligent decision model can be deployed in an intelligent decision system or a corresponding APP, which has an application layer, a service layer and a facility layer. The application layer is the application side of the intelligent decision model, which can call related decision services through the HTTP protocol. The service layer includes intelligent decision model management, scheduling execution and data and algorithm interface. The management part realizes the creation, editing, training, testing and release of the intelligent decision model and displays the details of each node (data) in the intelligent decision model. The intelligent decision model scheduling execution service reads the node configuration information (configured deviation interval, event, etc.) of the intelligent decision model through the HTTP protocol, and then calculates the information of each node according to the actual value, target value, deviation value, event and activity in turn, and reports the running result, i.e. wisdom data, to the application layer, while saving the data in the running process to the corresponding database in the facility layer. Thus, automatic decision is realized, the decision efficiency and accuracy are improved, and the complex and variable business scenarios are adapted.

[0044] Figure 8 is a schematic block diagram of an activity decision generation device 200 based on a decision model provided by an embodiment of the present application. As shown in Figure 8 Corresponding to the above activity decision generation method based on the decision model, the present application also provides an activity decision generation device based on the decision model. The activity decision generation device based on the decision model includes units for executing the above activity decision generation method based on the decision model. The device can be configured in a desktop computer, a tablet computer, a laptop computer, etc. terminal. Specifically, please refer to Figure 8 The activity decision generation device based on the decision model includes an acquisition unit 210, a matching unit 220, a calculation unit 230 and a determination unit 240.

[0045] The acquisition unit 210 is configured to acquire corresponding actual parameters and standard parameters from a preset data source according to business requirements.

[0046] In an embodiment, the acquisition unit 210 includes a generation unit, a comparison unit and an adjustment unit.

[0047] The generation unit is configured to generate a corresponding training execution activity according to training actual parameters and training standard parameters in a training data set. The comparison unit is configured to compare the training execution activity with a training label activity in the training data set to generate error information. The adjustment unit is configured to adjust corresponding model parameter values according to the error information to determine the intelligent decision model.

[0048] In an embodiment, the obtaining unit 210 comprises a judging unit and an adjusting sub-unit.

[0049] The judging unit is configured to judge whether the accuracy of the training execution activity generated by the adjusted intelligent decision-making model reaches a preset accuracy. The adjusting sub-unit is configured to continue adjusting the model parameter value according to the training data set if the accuracy does not reach the preset accuracy.

[0050] The matching unit 220 is configured to match the corresponding target deviation algorithm according to the actual parameter and the standard parameter.

[0051] In an embodiment, the matching unit 220 comprises a first screening unit and a second screening unit.

[0052] The first screening unit is configured to screen an initial deviation algorithm set within a preset similarity range according to the attribute information of the actual parameter and the standard parameter. The second screening unit is configured to screen the target deviation algorithm from the initial deviation algorithm set according to the data type of the actual parameter and the standard parameter.

[0053] The calculating unit 230 is configured to calculate the deviation value between the actual parameter and the standard parameter according to the target deviation algorithm.

[0054] The determining unit 240 is configured to match the corresponding event value according to the deviation value, and determine the corresponding execution activity according to the event value.

[0055] In an embodiment, the determining unit 240 comprises a matching unit and a first determining unit.

[0056] The matching unit is configured to match the deviation value in a preset deviation interval to obtain a target deviation interval. The first determining unit is configured to determine the corresponding event value according to the target deviation interval.

[0057] In the embodiment, the determining unit 240 comprises a type determining unit, a third screening unit and a fourth screening unit.

[0058] The type determining unit is configured to determine the type of the deviation value, wherein the deviation value comprises a positive deviation, a zero deviation and a negative deviation. The third screening unit is configured to screen the deviation value and its type in a plurality of preset first-level intervals to determine a target first-level interval. The fourth screening unit is configured to screen the deviation value in a plurality of preset second-level labels in the target first-level interval to determine the target deviation interval.

[0059] In this embodiment, the determining unit 240 includes an operation unit and an event determining unit.

[0060] The operation unit is configured to perform logical operation according to the deviation value to obtain an operation result. The event determining unit is configured to determine the event value corresponding to the target deviation interval according to the operation result.

[0061] It should be noted that the specific implementation process of the activity decision generation device 200 and each unit based on the decision model can be clearly understood by those skilled in the art, which can be referred to the corresponding description in the foregoing method embodiments. For the convenience and brevity of description, it will not be repeated here.

[0062] The activity decision generation device based on the decision model can be realized in the form of a computer program, which can run on a computer device as shown in the computer device. Figure 9

[0063] Please refer to Figure 9 , Figure 9 is a schematic block diagram of a computer device provided by an embodiment of the present application. The computer device 500 can be a terminal or a server, wherein the terminal can be a smart phone, a tablet computer, a notebook computer, a desktop computer, a personal digital assistant, a wearable device, and an electronic device with a communication function. The server can be a stand-alone server or a server cluster composed of multiple servers.

[0064] Referring to Figure 9 , the computer device 500 includes a processor 502, a memory, and a network interface 505 connected through a system bus 501, wherein the memory can include a non-volatile storage medium 503 and an internal memory 504.

[0065] The non-volatile storage medium 503 can store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions, which when executed, can cause the processor 502 to perform an activity decision generation method based on a decision model.

[0066] The processor 502 is configured to provide computing and control capabilities to support the operation of the entire computer device 500.

[0067] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503, which when executed by the processor 502, can cause the processor 502 to perform an activity decision generation method based on a decision model.

[0068] ​The network interface 505 is configured to perform network communication with other devices. Those skilled in the art can understand that Figure 9 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device 500 to which the scheme of the present application is applied. The specific computer device 500 can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0069] The processor 502 is configured to run the computer program 5032 stored in the memory to implement the steps of the above method.

[0070] It should be understood that, in the embodiments of the present application, the processor 502 can be a central processing unit (CPU), and the processor 502 can also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0071] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiments of the method can be completed by a computer program instructing related hardware. The computer program includes program instructions, and the computer program can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the above-mentioned embodiments of the method.

[0072] Therefore, the present application also provides a storage medium. The storage medium can be a computer-readable storage medium. The storage medium stores a computer program, wherein the computer program includes program instructions. The program instructions are executed by the processor to make the processor perform the steps of the above method.

[0073] The storage medium can be a U disk, a mobile hard disk, a read-only memory (ROM), a magnetic disk or an optical disk, and various computer-readable storage media that can store program codes.

[0074] Those skilled in the art can understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized in electronic hardware, computer software or a combination of both. In order to clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been described in general terms in the above description. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0075] In several embodiments provided by the present application, it should be understood that the disclosed apparatus and method can be implemented in other ways. For example, the apparatus embodiments described above are merely schematic. For example, the division of each unit is only a logical function division, and actual implementation can have another division manner. For example, a plurality of units or components can be combined or integrated into another system, or some features can be omitted or not executed.

[0076] The steps in the method embodiments of the present application can be adjusted, combined and deleted in sequence according to actual needs. The units in the apparatus embodiments of the present application can be combined, divided and deleted according to actual needs. In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit.

[0077] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a storage medium. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art, or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a terminal or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present application.

[0078] The above description is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical range disclosed by the present application, and these modifications or replacements should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for generating activity decisions based on a decision model, characterized in that: The method is applied to an intelligent decision-making model, and the method comprises: Obtain corresponding actual parameters and standard parameters from the preset data source according to business requirements; According to the target deviation algorithm corresponding to the matching of the actual parameters and the standard parameters; Calculating the deviation between the actual parameter and the standard parameter according to the target deviation algorithm; A corresponding event value is matched according to the deviation value, and a corresponding execution activity is determined according to the event value.

2. The method according to claim 1, characterized in that The step of matching the actual parameters with the standard parameters according to the corresponding target deviation algorithm includes: Screening out an initial deviation algorithm set within a preset similarity range based on the attribute information of the actual parameters and the standard parameters; The target deviation algorithm is selected from the initial deviation algorithm set according to the data types of the actual parameter and the standard parameter.

3. The method according to claim 1, characterized in that The step of matching the corresponding event value according to the deviation value includes: Matching the deviation value in a preset deviation range to obtain a target deviation range; A corresponding event value is determined according to the target deviation interval.

4. The method according to claim 3, characterized in that The step of matching the deviation value within the preset deviation range to obtain a target deviation range includes: Determining the type of the deviation value, wherein the deviation value includes positive deviation, zero deviation, and negative deviation; Screening a number of preset first-level intervals according to the deviation value and its type to determine a target first-level interval; The target deviation interval is determined by screening a plurality of preset secondary tags in the target primary interval according to the deviation value.

5. The method according to claim 3, characterized in that The step of determining the corresponding event value according to the target deviation interval includes: Performing a logical operation based on the deviation value to obtain an operation result; The event value corresponding to the target deviation interval is determined according to the calculation result.

6. The method according to claim 1, characterized in that Before the step of obtaining corresponding actual parameters and standard parameters from a preset data source according to business requirements, the following steps are included: Generate corresponding training execution activities based on the actual training parameters and standard training parameters in the training data set; comparing the training execution activity with the training label activity in the training data set and generating error information; The corresponding model parameter values ​​are adjusted according to the error information to determine the intelligent decision-making model.

7. The method according to claim 6, characterized in that The step of adjusting corresponding model parameter values ​​according to the error information to determine the intelligent decision model includes: Determine whether the accuracy of the training execution activities generated by the adjusted intelligent decision-making model reaches the preset accuracy rate; If not, continue to adjust the model parameter value according to the training data set.

8. An activity decision generating device based on a decision model, characterized in that: The device is applied to an intelligent decision-making model, and the device comprises: An acquisition unit, used to obtain corresponding actual parameters and standard parameters from a preset data source according to business requirements; A matching unit, configured to match a corresponding target deviation algorithm according to the actual parameters and the standard parameters; a calculation unit, configured to calculate a deviation value between the actual parameter and the standard parameter according to the target deviation algorithm; A determination unit is configured to match a corresponding event value according to the deviation value, and determine a corresponding execution activity according to the event value.

9. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the method according to any one of claims 1 to 7 when executing the computer program.

10. A storage medium, characterized in that: The storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the method according to any one of claims 1 to 7 can be implemented.

Images