Data monitoring method and device based on intelligent analysis, medium and equipment

Abstract

The invention relates to a data monitoring method and device based on intelligent analysis, a medium and equipment. The method comprises the steps of inputting index vectors corresponding to M index data into an association prediction model, outputting M prediction vectors, determining a first training loss according to the M index vectors, the M prediction vectors, a reference association matrix and a weight matrix corresponding to the association prediction model, further updating the weight matrix to obtain a target association matrix, and determining a second training loss according to the target association matrix. Determining target associated data corresponding to each piece of index data, monitoring the target associated data of the index data while monitoring the index data according to the user instruction, and supervising the prior information in the reference association matrix to be not lost through the first training loss; the association relationship among the index data can be found from the index vectors obtained from a plurality of index data in actual monitoring, so that the comprehensiveness and completeness of data monitoring are improved.

Description

Technical Field

[0001] The present invention is applicable to the field of data monitoring, and in particular relates to a data monitoring method, device, medium and equipment based on intelligent analysis. Background Art

[0002] Currently, in traditional data analysis applications, enterprises regularly check and monitor various indicator data and manually browse data along predefined query paths. However, only referring to predefined technical indicators and performance indicators will limit the perspective of data monitoring, and will miss other indicators that are related to the displayed indicators but not predefined, resulting in poor comprehensiveness and completeness of data monitoring.

[0003] Therefore, how to improve the comprehensiveness and completeness of data monitoring has become an urgent problem to be solved. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a data monitoring method, device, medium and equipment based on intelligent analysis to solve the problem.

[0005] In a first aspect, a data monitoring method based on intelligent analysis is provided, the method comprising: Get the indicator vectors corresponding to M indicator data respectively, where M is an integer greater than zero; Inputting M indicator vectors into a preset association prediction model, outputting prediction vectors corresponding to the M indicator data respectively, wherein the initial value of the weight matrix corresponding to the association prediction model is a preset reference association matrix; Determining a first training loss according to the M indicator vectors, the M prediction vectors, the reference association matrix, and a weight matrix corresponding to the association prediction model; According to the first training loss, a weight matrix corresponding to the association prediction model is updated until the first training loss converges to obtain a target association matrix; According to the target association matrix, determining the target association data corresponding to each indicator data; When any indicator data is monitored according to a user instruction, target associated data of the indicator data is monitored simultaneously.

[0006] In a second aspect, a data monitoring device based on intelligent analysis is provided, the device comprising: A vector acquisition module is used to obtain the indicator vectors corresponding to M indicator data, where M is an integer greater than zero; A vector prediction module, used for inputting M indicator vectors into a preset association prediction model, and outputting prediction vectors corresponding to the M indicator data respectively, wherein the initial value of the weight matrix corresponding to the association prediction model is a preset reference association matrix; A loss calculation module, used to determine a first training loss according to the M indicator vectors, the M prediction vectors, the reference association matrix and a weight matrix corresponding to the association prediction model; A weight updating module, used for updating the weight matrix corresponding to the association prediction model according to the first training loss until the first training loss converges to obtain a target association matrix; An association determination module, used to determine the target association data corresponding to each indicator data according to the target association matrix; The data monitoring module is used to monitor target associated data of any indicator data at the same time when monitoring the indicator data according to user instructions.

[0007] In a third aspect, an embodiment of the present invention provides a computer device, which includes a processor, a memory, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, it implements the data monitoring method of the first aspect.

[0008] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the data monitoring method of the first aspect is implemented.

[0009] Compared with the prior art, the present invention has the following beneficial effects: Target associated data that is not predefined in the indicator data is mined in the form of an associated prediction model, and the prior information in the reference association matrix is ​​not lost through the first training loss supervision. The association relationship between the indicator data can be found from the indicator vectors obtained from several indicator data in actual monitoring, thereby improving the comprehensiveness and completeness of data monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative labor. Figure 1 This is a schematic diagram of an application environment of a data monitoring method based on intelligent analysis provided by Embodiment 1 of the present invention; Figure 2 It is a flowchart of a data monitoring method based on intelligent analysis provided by Embodiment 1 of the present invention; Figure 3 It is a structural schematic diagram of a data monitoring device based on intelligent analysis provided by Embodiment 2 of the present invention; Figure 4 It is a structural schematic diagram of a computer device corresponding to a data monitoring method based on intelligent analysis provided in Embodiment 3 of the present invention; DETAILED DESCRIPTION

[0011] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0012] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0013] It should also be understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0014] As used in the present specification and the appended claims, the term "if" may be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" may be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]", depending on the context.

[0015] In addition, in the description of the present specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0016] References to "one embodiment" or "some embodiments" etc. described in the present specification mean that one or more embodiments of the present invention include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0017] The embodiments of the present invention can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0018] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0019] It should be understood that the order of execution of the steps in the following embodiments does not imply a precedence of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0020] In order to illustrate the technical solution of the present invention, specific embodiments are provided below for illustration.

[0021] A data monitoring method based on intelligent analysis provided by the first embodiment of the present invention can be applied in the following aspects: Figure 1In the application environment, the data input layer can provide the data monitoring method with monitoring data corresponding to each indicator data at each preset time point, that is, provide indicator vectors corresponding to each indicator data, and provide reference association relationships between each indicator data, that is, reference association matrix. This embodiment 1 can be applied to a database to perform data analysis, determine target association data for each indicator data, and provide target association relationships between indicator data for the system monitoring service in the service platform. When the user interacts with the application layer to perform the data display function, the target association data of the indicator data specified by the user can be displayed at the same time, which improves the comprehensiveness of data monitoring and provides users with a better data display experience.

[0022] See also Figure 2 , is a flow chart of a data monitoring method based on intelligent analysis provided by the first embodiment of the present invention. The above data monitoring method can be applied to Figure 1 The database in the data input layer obtains the indicator vectors corresponding to each indicator data and the reference association matrix. The database deploys an association prediction model to analyze the indicator vectors and determine the target association data corresponding to each indicator data, thereby providing a reference for the system monitoring service in the service platform, so that when the user interacts with the application layer to perform the data display function, the specified indicator data and its target association data can be displayed simultaneously. Figure 2 As shown, the data monitoring method may include the following steps: Step S201, obtaining indicator vectors corresponding to M indicator data respectively.

[0023] Wherein, M is an integer greater than zero, the indicator data may refer to the indicator to be monitored, and the indicator vector may be a vector formed by monitoring data obtained by monitoring the corresponding indicator data at several preset time points.

[0024] Specifically, the indicator vector can be obtained from the monitoring data stream of the corresponding indicator data in the form of a sliding window. The size, step size and starting position of the sliding window are all default values ​​in this embodiment, and in this embodiment, the parameters of the sliding windows corresponding to each indicator vector are consistent, for example, the sizes are all preset sizes, the step sizes are all 1, and the starting positions are all monitoring data corresponding to a specific preset time point.

[0025] Step S202: input the M indicator vectors into a preset association prediction model, and output prediction vectors corresponding to the M indicator data.

[0026] Among them, the initial value of the weight matrix corresponding to the association prediction model is a preset reference association matrix. The association prediction model can be used to predict a specific indicator vector based on a partial indicator vector. The prediction vector can refer to the prediction result of a specific indicator vector based on other indicator vectors.

[0027] Specifically, in this embodiment, the association prediction model uses a fully connected network connection method, but it does not use a fully connected network to perform tasks such as feature extraction like the prior art, but instead uses a fully connected network connection method to perform correlation analysis between indicator data.

[0028] The input layer of the association prediction model may include M first neurons and M second neurons, each first neuron uniquely corresponds to an indicator vector as input, the outputs of the M first neurons are connected to the M second neurons, the second neurons correspond to the prediction vector, and for any first neuron and any second neuron, the connection relationship between the two corresponds to a weight value, which in this embodiment represents the degree of association between the indicator data corresponding to the first neuron and the indicator data corresponding to the second neuron. Moreover, since the association prediction model in this embodiment is based on a fully connected network, the weight value can actually be updated through the training of the fully connected network. That is, this embodiment uses the existing fully connected network to implement the indicator association analysis task, which is not a conventional application method of the fully connected network.

[0029] Optionally, in the reference association matrix, the elements corresponding to two indicator data having a preset association relationship are first preset values, and the other elements are second preset values.

[0030] The preset association relationship may be defined by the implementer, the first preset value may be 1, and the second preset value may be 0.

[0031] Specifically, if any two indicator data correspond to an element in the weight matrix of the association prediction model, the weight matrix of the association prediction model can be used to characterize the association relationship between the M indicator data. The element corresponding to the two indicator data with a preset association relationship in the reference association matrix is ​​the first preset value, and the other elements are the second preset values. In practice, the implementer sets the element corresponding to the two indicator data with a known association relationship to the first preset value.

[0032] It should be noted that, in the reference association matrix, the elements of the main diagonal are all second preset values, that is, the elements of the first preset value need to correspond to two different indicator data. The reason is that this embodiment takes into account factors such as the complexity and reusability of model construction, and selects the connection method of the fully connected network to construct the association prediction model. However, when mining the correlation of indicator data, it is not hoped that the indicator data itself will contribute to the prediction of the indicator data, but it is hoped that specific indicator data can be predicted through other indicator data and reflected in the reference association matrix, that is, the elements of the main diagonal are all second preset values.

[0033] This embodiment uses the reference association matrix as the initial value of the weight matrix of the association prediction model, which can provide known prior information for the indicator data association analysis, such as the known association relationship between the indicator data, and introduces the reference association matrix in the calculation of the subsequent loss function, so that the training process will not lose the existing prior information, and avoid being affected by the specific indicator data itself when predicting, resulting in the training results of the association prediction model being unable to characterize the association relationship between the indicator data.

[0034] Step S203, determining a first training loss according to the M indicator vectors, the M prediction vectors, the reference association matrix, and the weight matrix corresponding to the association prediction model.

[0035] Among them, the first training loss can be used to supervise the training of the weight matrix corresponding to the association prediction model.

[0036] Optionally, determining the first training loss according to the M indicator vectors, the M prediction vectors, the reference association matrix, and the weight matrix corresponding to the association prediction model includes: Determine a first training sub-loss according to the M indicator vectors and the M prediction vectors; Determining a second training sub-loss according to the reference association matrix and the weight matrix corresponding to the association prediction model; The sum of the first training sub-loss and the second training sub-loss is taken as the first training loss.

[0037] Among them, the first training sub-loss can be used to characterize the difference between the predicted value of the association prediction model for each indicator data, that is, the prediction vector, and the true value of the indicator data, that is, the indicator vector, so that the supervised association prediction model can predict specific indicator data based on other indicator data.

[0038] The second training sub-loss can be used to supervise that the prior information in the reference association matrix is ​​not lost, thereby avoiding the loss of the prior information pre-introduced by the implementer as the weight matrix corresponding to the association prediction model is updated during the training process, thereby ensuring that the update of the weight matrix corresponding to the association prediction model is better than or equal to the existing prior knowledge.

[0039] Optionally, determining the first training sub-loss according to the M indicator vectors and the M prediction vectors includes: For any indicator data, the Euclidean distance between the indicator vector and the prediction vector respectively corresponding to the indicator data is calculated to obtain a first distance value of the corresponding indicator data; A first training sub-loss is determined according to the first distance values ​​respectively corresponding to the M indicator data.

[0040] The first distance value may represent the difference between the indicator vector and the prediction vector of the corresponding indicator data.

[0041] Optionally, determining the first training sub-loss according to the first distance values ​​respectively corresponding to the M indicator data includes: For any indicator data, multiply the first distance value corresponding to the indicator data and the reference weight value corresponding to the indicator data to obtain a second distance value corresponding to the indicator data; Add the second distance values ​​corresponding to the M indicator data respectively, and use the added result as the distance loss; Add the reference weight values ​​corresponding to the M indicator data respectively, calculate the absolute value of the difference between the addition result and the third preset value, and take the product of the calculation result and the fourth preset value as the weight loss; The sum of distance loss and weight loss is used as the first training sub-loss.

[0042] Among them, in this embodiment, the initial value of the reference weight value of each indicator data is 1, and the third preset value is M. The purpose of setting the reference weight value is to avoid the failure of the training process to converge due to the emergence of independent indicator data. Independent indicator data can refer to indicator data that has no correlation with any other indicator data.

[0043] Specifically, the initial value of the fourth preset value is a maximum value, for example, set to 256. The maximum value can be determined by the implementer according to the empirical value of the distance loss, and the maximum value should be greater than the empirical value of the distance loss.

[0044] When only considering the distance loss and weight loss, after the training starts, since the initial value of the reference weight value of each indicator data is 1 and the third preset value is M, it can be known that the weight loss is 0. At this time, since the fourth preset value is larger, the training process pays more attention to the weight loss. In order to ensure that the weight loss continues to be 0, the reference weight value of each indicator data will not change during the training process. At this time, if there is independent indicator data, the distance loss will be difficult to converge, and accordingly, the first training sub-loss will also be difficult to converge. If the first training sub-loss has not converged after the preset rounds, the fourth preset value is exponentially decayed, that is, the fourth preset value is squared, and the fourth preset value in the weight loss is updated with the processed fourth preset value.

[0045] After the fourth preset value is updated, if the first training sub-loss has not converged after the preset rounds, it means that the training process is still too focused on the weight loss. The reason is that when the training process reduces the focus on weight loss, the reference weight value can be adjusted. In extreme cases, that is, when the fourth preset value is close to 0, the training process can completely ignore the value of the reference weight value. The training process can update the reference weight value to 0, and the corresponding distance loss must be 0. The first training sub-loss must converge to the product of the fourth preset value and M. Therefore, the fourth preset value is exponentially decayed again to update the weight loss with the fourth preset value after processing until the first training sub-loss converges.

[0046] It should be noted that since the exponential decay of the fourth preset value gradually decays with the iteration rounds, and the decay amplitude gradually decreases, the fourth preset value usually does not reach the situation where it is close to 0. In general, when the training process focuses on weight loss and distance loss at the same time, the training process will update the reference weight value corresponding to the independent indicator data to close to 0, thereby achieving the purpose of convergence of the first loss sub-loss with the minimum reference weight value update amplitude.

[0047] This embodiment adds a reference weight value to the first distance corresponding to each indicator data, and adds an additional weight loss according to the reference weight value, so that the training process can effectively eliminate the influence of independent indicator data, avoid the situation where the training process is difficult to converge, and improve the feasibility of the data monitoring method.

[0048] Optionally, determining the second training sub-loss according to the reference association matrix and the weight matrix corresponding to the association prediction model includes: Calculate the transposed matrix of the weight matrix, subtract the weight matrix and the transposed matrix, and take the sum of the absolute values ​​of all elements in the difference result as the first matrix loss; Calculate the trace of the weight matrix and use the result as the second matrix loss; Multiply the weight matrix and the reference association matrix point by point, subtract the result of the point-by-point multiplication from the reference association matrix, and take the sum of the absolute values ​​of all elements in the difference result as the third matrix loss; The sum of the first matrix loss, the second matrix loss and the third matrix loss is taken as the second training sub-loss.

[0049] Among them, in the weight matrix, the element in the i-th row and the j-th column represents the correlation between the i-th indicator data and the j-th indicator data. Obviously, when the i-th indicator data and the j-th indicator data are correlated, the j-th indicator data and the i-th indicator data must also be correlated. Therefore, the weight matrix must be a symmetric matrix, and the diagonal elements of the weight matrix are all 0.

[0050] Specifically, according to the properties of symmetric matrices, the transposed matrix of the weight matrix should be the same as itself, so the weight matrix can be supervised to be a symmetric matrix through the first matrix loss.

[0051] The trace of the weight matrix is ​​the sum of the diagonal elements of the weight matrix, so the diagonal elements of the weight matrix can be supervised to be 0 through the second matrix loss.

[0052] The values ​​corresponding to the elements in the reference association matrix only have the first preset value and the second preset value. The first preset value can be 1, and the second preset value can be 0. In order to ensure that the value corresponding to the element corresponding to the first preset value in the reference association matrix is ​​still 1 in the weight matrix, the weight matrix and the reference association matrix are multiplied point by point, and the point-by-point multiplication result is subtracted from the reference association matrix, and the sum of the absolute values ​​of all elements in the subtraction result is used as the third matrix loss, thereby ensuring that the training process does not lose the prior information of the reference association matrix.

[0053] In this embodiment, a first matrix loss, a second matrix loss and a third matrix loss are constructed to supervise that the prior information in the reference association matrix is ​​not lost, so as to avoid the loss of the prior information pre-introduced by the implementer as the weight matrix corresponding to the association prediction model is updated during the training process, thereby ensuring that the update of the weight matrix corresponding to the association prediction model is better than or equal to the existing prior knowledge.

[0054] Step S204: updating the weight matrix corresponding to the association prediction model according to the first training loss until the first training loss converges to obtain a target association matrix.

[0055] Among them, the values ​​corresponding to the elements in the target association matrix can be the second preset value and other values, and the second preset value can be 0. When any element is the second preset value, it means that there is no correlation between the two indicator data corresponding to the element, otherwise, it means that there is a correlation between the two indicator data corresponding to the element.

[0056] Step S205: determining the target associated data corresponding to each indicator data according to the target associated matrix.

[0057] The target-related data may refer to indicator data that is associated with the corresponding indicator data.

[0058] Optionally, according to the target association matrix, the target association data corresponding to each indicator data is determined, including: For any indicator data, determine the target association vector corresponding to the indicator data from the target association matrix; The indicator data corresponding to the elements in the target association vector that are not the second preset value are used as the target association data of the indicator data.

[0059] Among them, the second preset value can be 0, and the target association vector can refer to the row vector or column vector corresponding to the indicator data. Since the target association matrix is ​​a symmetric matrix, the row vector and column vector corresponding to the indicator data are transposed to each other.

[0060] Step S206, when any indicator data is monitored according to the user instruction, the target associated data of the indicator data is monitored at the same time.

[0061] Among them, the user can monitor specific indicator data through user instructions. At this time, the target-related data of the specific indicator data is monitored at the same time. In another embodiment, all indicator data can be monitored at the same time. The user can display the monitored specific indicator data through user instructions. At this time, the monitoring data corresponding to the target-related data of the specific indicator data is displayed at the same time, thereby improving the comprehensiveness and completeness of data monitoring and data display, and improving the user's data monitoring and data display experience.

[0062] In this embodiment, target associated data that is not predefined in the indicator data is mined in the form of an associated prediction model, and the prior information in the reference association matrix is ​​not lost through the first training loss supervision. The association relationship between the indicator data can be found from the indicator vectors obtained from several indicator data in actual monitoring, thereby improving the comprehensiveness and completeness of data monitoring.

[0063] Corresponding to the method of the above embodiment 1, Figure 3 A schematic diagram of the structure of a data monitoring device based on intelligent analysis provided by the second embodiment of the present invention is shown. The data monitoring device is applied to a database. The database obtains the indicator vectors corresponding to each indicator data and the reference association matrix from the data input layer. An association prediction model is deployed in the database to analyze the indicator vectors and determine the target association data corresponding to each indicator data, thereby providing a reference for the system monitoring service in the service platform, so that when the user interacts with the application layer to perform the data display function, the specified indicator data and its target association data can be displayed simultaneously. For the convenience of explanation, only the part related to the embodiment of the present invention is shown.

[0064] See also Figure 3 , the data monitoring device comprises: A vector acquisition module 31 is used to acquire indicator vectors corresponding to M indicator data, where M is an integer greater than zero; The vector prediction module 32 is used to input the M indicator vectors into a preset association prediction model, and output the prediction vectors corresponding to the M indicator data respectively, wherein the initial value of the weight matrix corresponding to the association prediction model is a preset reference association matrix; A loss calculation module 33, used to determine a first training loss according to the M indicator vectors, the M prediction vectors, the reference association matrix and the weight matrix corresponding to the association prediction model; A weight updating module 34 is used to update the weight matrix corresponding to the association prediction model according to the first training loss until the first training loss converges to obtain a target association matrix; The association determination module 35 is used to determine the target association data corresponding to each indicator data according to the target association matrix; The data monitoring module 36 is used to monitor the target associated data of the indicator data while monitoring any indicator data according to the user instruction.

[0065] Optionally, in the reference association matrix in the above-mentioned vector prediction module 32, the elements corresponding to two indicator data having a preset association relationship are first preset values, and the other elements are second preset values.

[0066] Optionally, the loss calculation module 33 includes: A first sub-loss submodule, used to determine a first training sub-loss according to the M indicator vectors and the M prediction vectors; A second sub-loss sub-module, used to determine a second training sub-loss according to a reference association matrix and a weight matrix corresponding to the association prediction model; The first loss determination submodule is configured to use the sum of the first training sub-loss and the second training sub-loss as the first training loss.

[0067] Optionally, the first sub-loss sub-module includes: A first distance unit is used to calculate the Euclidean distance between the indicator vector and the prediction vector respectively corresponding to any indicator data, and obtain a first distance value of the corresponding indicator data; The first sub-loss determining unit determines the first training sub-loss according to the first distance values ​​respectively corresponding to the M indicator data.

[0068] Optionally, the first sub-loss determination unit includes: A second distance subunit is used to multiply the first distance value corresponding to the indicator data and the reference weight value corresponding to the indicator data for any indicator data to obtain a second distance value corresponding to the indicator data; A distance loss subunit, used for adding the second distance values ​​corresponding to the M indicator data respectively, and taking the addition result as the distance loss; A weight loss subunit, used to add the reference weight values ​​corresponding to the M indicator data respectively, calculate the absolute value of the difference between the addition result and the third preset value, and use the product of the calculation result and the fourth preset value as the weight loss; The first sub-loss determination subunit is used to use the sum of the distance loss and the weight loss as the first training sub-loss.

[0069] Optionally, the second sub-loss sub-module includes: The first matrix loss submodule is used to calculate the transposed matrix of the weight matrix, subtract the weight matrix and the transposed matrix, and take the sum of the absolute values ​​of all elements in the difference result as the first matrix loss; The second matrix loss submodule is used to calculate the trace of the weight matrix and use the calculation result as the second matrix loss; The third matrix loss submodule is used to multiply the weight matrix and the reference association matrix point by point, and to make a difference between the point-by-point multiplication result and the reference association matrix, and to take the sum of the absolute values ​​of all elements in the difference result as the third matrix loss; The second sub-loss determination submodule is used to use the sum of the first matrix loss, the second matrix loss and the third matrix loss as the second training sub-loss.

[0070] Optionally, the association determination module 35 includes: The association vector determination submodule is used to determine the target association vector corresponding to the indicator data from the target association matrix for any indicator data; The associated data determination submodule is used to use the indicator data corresponding to the elements in the target associated vector that are not the second preset value as the target associated data of the indicator data.

[0071] It should be noted that the information interaction, execution process, etc. between the above-mentioned modules, sub-modules, units, and sub-units are based on the same concept as the embodiment of the method of the present invention. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0072] Figure 4 A schematic diagram of the structure of a computer device corresponding to a data monitoring method based on intelligent analysis provided in Embodiment 3 of the present invention. Figure 4 As shown, the computer device of this embodiment includes: at least one processor ( Figure 4 Only one is shown in the figure), a memory, and a computer program stored in the memory and executable on at least one processor, wherein when the processor executes the computer program, the steps in any of the above-mentioned data monitoring method embodiments are implemented.

[0073] The computer device may include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that Figure 4 This is merely an example of a computer device and does not constitute a limitation on the computer device. The computer device may include more or fewer components than those shown in the figure, or a combination of certain components, or different components. For example, it may also include a network interface, a display screen, and an input device.

[0074] The processor may be a CPU, or other general-purpose processors, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0075] The memory includes a readable storage medium, an internal memory, etc., wherein the internal memory may be the memory of a computer device, and the internal memory provides an environment for the operation of the operating system and computer-readable instructions in the readable storage medium. The readable storage medium may be a hard disk of a computer device, and in other embodiments, it may also be an external storage device of the computer device, for example, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the computer device. Further, the memory may also include both an internal storage unit of the computer device and an external storage device. The memory is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of a computer program, etc. The memory may also be used to temporarily store data that has been output or is to be output.

[0076] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention. The specific working process of the units and modules in the above-mentioned device can refer to the corresponding process in the above-mentioned method embodiment, which will not be repeated here. If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned method embodiment can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device capable of carrying computer program code, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0077] The present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by a computer program product. When the computer program product runs on a computer device, the computer device can implement the steps in the above-mentioned method embodiment when executing.

[0078] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0079] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0080] In the embodiments provided by the present invention, it should be understood that the disclosed devices / computer equipment and methods can be implemented in other ways. For example, the device / computer equipment embodiments described above are only schematic, for example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0081] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0082] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. A data monitoring method based on intelligent analysis, characterized in that: The data monitoring method comprises: Get the indicator vectors corresponding to M indicator data respectively, where M is an integer greater than zero; Inputting M indicator vectors into a preset association prediction model, outputting prediction vectors corresponding to the M indicator data respectively, wherein the initial value of the weight matrix corresponding to the association prediction model is a preset reference association matrix; Determining a first training loss according to the M indicator vectors, the M prediction vectors, the reference association matrix, and a weight matrix corresponding to the association prediction model; According to the first training loss, a weight matrix corresponding to the association prediction model is updated until the first training loss converges to obtain a target association matrix; According to the target association matrix, determining the target association data corresponding to each indicator data; When any indicator data is monitored according to a user instruction, target associated data of the indicator data is monitored simultaneously.

2. The data monitoring method according to claim 1, characterized in that: In the reference association matrix, the elements corresponding to two indicator data having a preset association relationship are first preset values, and the other elements are second preset values.

3. The data monitoring method according to claim 1, characterized in that: The determining a first training loss according to the M indicator vectors, the M prediction vectors, the reference association matrix, and a weight matrix corresponding to the association prediction model includes: Determining a first training sub-loss according to the M indicator vectors and the M prediction vectors; Determining a second training sub-loss according to the reference association matrix and a weight matrix corresponding to the association prediction model; The sum of the first training sub-loss and the second training sub-loss is taken as the first training loss.

4. The data monitoring method according to claim 3, characterized in that: Determining a first training sub-loss according to the M indicator vectors and the M prediction vectors includes: For any indicator data, calculating the Euclidean distance between the indicator vector and the prediction vector respectively corresponding to the indicator data, and obtaining a first distance value corresponding to the indicator data; The first training sub-loss is determined according to the first distance values ​​respectively corresponding to the M indicator data.

5. The data monitoring method according to claim 4, characterized in that: The determining the first training sub-loss according to the first distance values ​​respectively corresponding to the M indicator data includes: For any indicator data, multiplying a first distance value corresponding to the indicator data and a reference weight value corresponding to the indicator data to obtain a second distance value corresponding to the indicator data; Adding the second distance values ​​respectively corresponding to the M indicator data, and taking the addition result as the distance loss; Add the reference weight values ​​corresponding to the M indicator data respectively, calculate the absolute value of the difference between the addition result and the third preset value, and take the product of the calculation result and the fourth preset value as the weight loss; The sum of the distance loss and the weight loss is taken as the first training sub-loss.

6. The data monitoring method according to claim 3, characterized in that: The determining the second training sub-loss according to the reference association matrix and the weight matrix corresponding to the association prediction model includes: Calculate and obtain the transposed matrix of the weight matrix, subtract the weight matrix from the transposed matrix, and take the sum of the absolute values ​​of all elements in the subtracted result as the first matrix loss; Calculate the trace of the weight matrix and use the calculated result as the second matrix loss; Multiplying the weight matrix and the reference association matrix point by point, subtracting the point by point multiplication result from the reference association matrix, and taking the sum of the absolute values ​​of all elements in the subtraction result as the third matrix loss; The sum of the first matrix loss, the second matrix loss and the third matrix loss is taken as the second training sub-loss.

7. The data monitoring method according to claims 1 to 6, characterized in that: Determining the target association data corresponding to each indicator data according to the target association matrix includes: For any indicator data, determining a target association vector corresponding to the indicator data from the target association matrix; The indicator data corresponding to the elements in the target association vector that are not the second preset value are used as the target association data of the indicator data.

8. A data monitoring device based on intelligent analysis, characterized in that: The data monitoring device comprises: A vector acquisition module is used to obtain the indicator vectors corresponding to M indicator data, where M is an integer greater than zero; A vector prediction module, used for inputting M indicator vectors into a preset association prediction model, and outputting prediction vectors corresponding to the M indicator data respectively, wherein the initial value of the weight matrix corresponding to the association prediction model is a preset reference association matrix; A loss calculation module, used to determine a first training loss according to the M indicator vectors, the M prediction vectors, the reference association matrix and a weight matrix corresponding to the association prediction model; A weight updating module, used for updating the weight matrix corresponding to the association prediction model according to the first training loss until the first training loss converges to obtain a target association matrix; An association determination module, used to determine the target association data corresponding to each indicator data according to the target association matrix; The data monitoring module is used to monitor target associated data of any indicator data at the same time when monitoring the indicator data according to user instructions.

9. A computer device, characterized in that: The computer device includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the data monitoring method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the data monitoring method according to any one of claims 1 to 7 is implemented.

Images