WO2023093070A1

WO2023093070A1 - Intelligent city network resource-oriented correlation analysis method and device

Info

Publication number: WO2023093070A1
Application number: PCT/CN2022/104835
Authority: WO
Inventors: 杨杨; 高志鹏; 葛忠迪; 曲珍莹; 胡皓; 吕睿; 何晔辰; 范成文; 赵斌男; 郭延鹏
Original assignee: 北京邮电大学
Priority date: 2021-11-24
Filing date: 2022-07-11
Publication date: 2023-06-01
Also published as: CN114186168A

Abstract

The present application provides an intelligent city network resource-oriented correlation analysis method and device. The method comprises: obtaining a plurality of groups of different attribute variables of an intelligent city network, and obtaining the optimal correlation coefficient of the attribute variables on the basis of canonical correlation analysis; mapping the attribute variables corresponding to the optimal correlation coefficient to a subspace on the basis of a multi-kernel model to obtain a network operation data feature vector, wherein the multi-kernel model is established by performing linear combination on the basis of multiple types of kernel functions; and calculating a distance of the network operation data feature vector on the basis of Euclidean distance measurement, and obtaining the correlation of the network operation data feature vector according to the distance and a weight of the linear combination of the kernel functions.

Description

Correlation analysis method and device for smart city network resources

Cross References to Related Applications

This application claims the priority of the Chinese patent application with the application number 2021114049511 filed on November 24, 2021, and the title of the invention is "Correlation analysis method and device for smart city network resources", which is incorporated by reference in its entirety Apply.

technical field

The present application relates to the technical field of information processing, in particular to a correlation analysis method and device for smart city network resources.

Background technique

With the continuous innovation and progress of information technology, the construction of smart cities has been gradually put on the agenda, which has become the focus of current network facility recommendations and future urban planning and construction. Huge urban information construction projects require strong network infrastructure support. Complex network structures will generate massive amounts of network data. Most of the attribute data generated by network operation are high-dimensional nonlinear, which includes network traffic size, network transmission methods, network Addresses and other relevant important information also have more secretive network attributes. Therefore, how to extract key relevant feature data for analysis and research from massive network data has become the focus of current research, which is also one of the key issues that need to be broken through in network construction in smart city construction.

The method for extracting linear features in the prior art often uses Canonical Correlation Analysis (CCA) to analyze the correlation of network data. Canonical Correlation Analysis (CCA) can learn to make two sets of heterogeneous data The common subspace where the linear correlation is maximized, and the mapping of heterogeneous data to the common subspace is completed. But it is difficult to extract effective and key features only through the linear mapping of canonical correlation analysis.

Therefore, this subject urgently needs to solve the problem that it is difficult to extract nonlinear features from network data in the prior art.

Contents of the invention

This application provides a correlation analysis method and device for smart city network resources, which is used to solve the defect that it is difficult to extract nonlinear features from network data in the prior art, and realize linear extraction of network data to perform correlation analysis.

This application provides a correlation analysis method for smart city network resources, including:

Obtain multiple sets of different attribute variables of the smart city network, and obtain the optimal correlation coefficient of the attribute variables based on canonical correlation analysis;

Mapping the attribute variable corresponding to the optimal correlation coefficient to a subspace based on a multi-kernel model to obtain a feature vector of network operation data; wherein the multi-kernel model is established based on a linear combination of multiple kernel functions;

The distance of the feature vector of the network operation data is calculated based on the Euclidean distance measure, and the correlation of the feature vector of the network operation data is obtained according to the weight of the linear combination of the distance and the kernel function.

According to a correlation analysis method for smart city network resources provided by the present application, the multi-kernel model is established based on a linear combination of various kernel functions, specifically including:

Obtaining multiple types of kernel functions, linearly combining the kernel functions according to different weights, and obtaining a multi-kernel model after weight accumulation and linear combination of various kernel functions;

Wherein, the kernel function includes at least one of a polynomial kernel function, an exponential kernel function, a Gaussian kernel function and a linear kernel function.

According to a correlation analysis method for smart city network resources provided by the present application, the two sets of attribute variables corresponding to the optimal correlation coefficient are mapped to subspaces based on the multi-core model to obtain network operation data feature vectors, specifically including :

Based on each kernel function in the multi-kernel model, two sets of feature vectors corresponding to the two sets of attribute variables corresponding to the optimal correlation coefficient are mapped to the subspace;

The calculation of the distance of the feature vector of the network operation data based on the Euclidean distance metric specifically includes:

The distance between two sets of feature vectors corresponding to each kernel function is calculated based on the Euclidean distance metric.

According to a correlation analysis method for smart city network resources provided by the present application, the correlation of the network operation data feature vector is obtained according to the weight of the distance and the kernel function, specifically including:

According to the distance between the weight of each kernel function in the multi-kernel model and the two groups of eigenvectors corresponding to each kernel function, the correlation value of each kernel function is obtained;

The correlation values of each kernel function are summed to obtain the correlation of the feature vectors of the network operation data.

According to a correlation analysis method for smart city network resources provided by the present application, the optimal correlation coefficient of the attribute variable is obtained based on the canonical correlation analysis, which specifically includes:

Based on the canonical correlation analysis, the coefficient vector of the linear combination of the attribute variables and the typical variables after the linear combination are obtained;

Obtaining the correlation coefficient of the typical variable, adjusting the coefficient vector to make the correlation coefficient an optimal value, and obtaining the optimal correlation coefficient.

According to a correlation analysis method for smart city network resources provided by this application, it also includes:

Combining multiple sets of attribute variables in pairs, and obtaining the correlation coefficient corresponding to each combination;

The combination with the largest correlation coefficient is confirmed as the two groups of attribute variables corresponding to the optimal correlation coefficient.

The present application also provides a correlation analysis device for smart city network resources, including:

The optimal correlation coefficient acquisition module is used to obtain multiple sets of different attribute variables of the smart city network, and obtain the optimal correlation coefficient of the attribute variables based on canonical correlation analysis;

The feature vector acquisition module is used to map the attribute variable corresponding to the optimal correlation coefficient to the subspace based on the multi-kernel model to obtain the feature vector of the network operation data; wherein the multi-kernel model is established based on a linear combination of various kernel functions ;

The correlation analysis module is used to calculate the distance of the feature vector of the network operation data based on the Euclidean distance measure, and obtain the correlation of the feature vector of the network operation data according to the weight of the linear combination of the distance and the kernel function.

According to a correlation analysis device for smart city network resources provided by the present application, the feature vector acquisition module is specifically used to: acquire various types of kernel functions, and linearly combine the kernel functions according to different weights to obtain various Kernel function weight accumulation and multi-kernel model after linear combination;

According to a correlation analysis device for smart city network resources provided by the present application, the feature vector acquisition module is also used to: obtain two sets of attribute variables corresponding to the optimal correlation coefficient based on each kernel function in the multi-kernel model Two sets of eigenvectors mapped to the subspace;

The correlation analysis module is specifically used for: calculating the distance between two groups of feature vectors corresponding to each kernel function based on the Euclidean distance measure.

According to a correlation analysis device for smart city network resources provided by the present application, the correlation analysis module is also used for: according to the weight of each kernel function in the multi-kernel model, two groups corresponding to each kernel function The distance of the eigenvector to get the correlation value of each kernel function;

According to a correlation analysis device for smart city network resources provided by the present application, the optimal correlation coefficient acquisition module is specifically used to: obtain the coefficient vector of the linear combination of the attribute variables and the typical linear combination based on the typical correlation analysis. variable;

According to the correlation analysis device for smart city network resources provided in the present application, the correlation analysis device for smart city network resources provided in the present application further includes the following modules:

Combination module: combine multiple sets of attribute variables in pairs, and obtain the correlation coefficient corresponding to each combination;

Confirmation module: confirm the combination with the largest correlation coefficient as the two groups of attribute variables corresponding to the optimal correlation coefficient.

The present application also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and operable on the processor. When the processor executes the program, it realizes the smart city-oriented The steps of the correlation analysis method of network resources.

The present application also provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of any one of the aforementioned smart city network resource-oriented correlation analysis methods are implemented. .

The present application also provides a computer program product, including a computer program. When the computer program is executed by a processor, the steps of any one of the above-mentioned smart city network resource-oriented correlation analysis methods are implemented.

The present application also provides a chip, the chip includes a processor and a communication interface, the communication interface is coupled to the processor, and the processor is used to run programs or instructions to realize the smart city-oriented Steps of correlation analysis of network resources.

The correlation analysis method and device for smart city network resources provided by this application obtains the optimal correlation coefficient of the attribute variable of the smart city network through typical correlation analysis, and maps the attribute variable corresponding to the optimal correlation coefficient to subspace. The present application can process nonlinear network data by combining multiple kernel functions with canonical correlation analysis, and obtain more accurate correlation of network data.

Description of drawings

In order to more clearly illustrate the technical solutions in this application or the prior art, the accompanying drawings that need to be used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings in the following description are the present For some embodiments of the application, those skilled in the art can also obtain other drawings based on these drawings without creative work.

Fig. 1 is one of the schematic flow charts of the correlation analysis method for smart city network resources provided by the present application;

Fig. 2 is the second schematic flow diagram of the correlation analysis method for smart city network resources provided by the present application;

Fig. 3 is one of the schematic structural diagrams of the correlation analysis device for smart city network resources provided by the present application;

FIG. 4 is a schematic structural diagram of an electronic device provided by the present application.

Detailed ways

In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be clearly and completely described below in conjunction with the accompanying drawings in this application. Obviously, the described embodiments are part of the embodiments of this application , but not all examples. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of this application.

The following describes the correlation analysis method for smart city network resources provided by the present application with reference to FIGS. 1-2 .

Referring to Figure 1, the application provides a correlation analysis method for smart city network resources, including the following steps:

Step 110: Obtain multiple sets of different attribute variables of the smart city network, and obtain the optimal correlation coefficient of the attribute variables based on canonical correlation analysis;

In the actual urban information construction project, strong network infrastructure support is needed. Due to the complexity of the smart city network structure, massive network data will be generated. Most of the attribute data generated by network operation is high-dimensional nonlinear, which contains important information such as network traffic size, network propagation method, and network address, as well as relatively secret network attributes.

In this embodiment, the attribute variables in the smart city network are divided into multiple groups, and the optimal correlation coefficients of these attribute variables are obtained according to the canonical correlation analysis algorithm.

Canonical Correlation Analysis Algorithm (CCA, Canonical Correlation Analysis) is a multivariate statistical analysis method that uses the correlation between comprehensive variable pairs to reflect the overall correlation between two groups of indicators. The basic principle is: In order to grasp the correlation between the two groups of indicators as a whole, two representative comprehensive variables are extracted from the two groups of variables, and the correlation between the two comprehensive variables is used to reflect the correlation between the two groups of indicators. overall correlation between them.

In this embodiment, two sets of attribute variables with the greatest correlation are selected from multiple sets of attribute variables, and an optimal correlation coefficient is obtained from these two sets of attribute variables. Through the canonical correlation analysis algorithm, the correlation value between the two attribute variables is maximized.

Step 120: Map the attribute variable corresponding to the optimal correlation coefficient to a subspace based on a multi-kernel model to obtain a feature vector of network operation data; wherein, the multi-kernel model is established based on a linear combination of various kernel functions;

Specifically, in this embodiment, the two sets of attribute variables corresponding to the optimal correlation coefficient are projected into the high-dimensional space through the multi-kernel model to obtain the network operation data feature vector.

Among them, the multi-kernel model is to construct a linear combination of these kernel functions according to a variety of selected kernel functions, assign weights to each kernel function, and form a linear combination according to the weights to obtain a multi-kernel model. The multi-kernel model in this example should have at least one type of kernel function.

Step 130: Calculate the distance of the feature vectors of the network operating data based on the Euclidean distance metric, and obtain the correlation of the feature vectors of the network operating data according to the weight of the linear combination of the distance and the kernel function.

The Euclidean distance metric, also known as the Euclidean metric, refers to the real distance between two points in the m-dimensional space, or the natural length of the vector (that is, the distance from the point to the origin). The Euclidean distance in 2D and 3D space is the actual distance between two points.

In this embodiment, the distance between two groups of mapped network operating data feature vectors is calculated by the Euclidean distance measure, and the network is calculated according to the calculated distance and the weight of each kernel function in the multi-kernel model Run the correlation of data feature vectors, that is, the correlation of two groups of attribute variables before mapping.

The correlation analysis method for smart city network resources provided by the embodiment of the present application obtains the optimal correlation coefficient of the attribute variable of the smart city network through typical correlation analysis, and maps the attribute variable corresponding to the optimal correlation coefficient to subspace. The present application can process nonlinear network data by combining multiple kernel functions with canonical correlation analysis, and obtain more accurate correlation of network data.

Based on the above-mentioned embodiments, the multi-kernel model is established based on a linear combination of various kernel functions, specifically including:

In this embodiment, the multi-kernel model is based on multiple selected kernel functions, constructing a linear combination of these kernel functions, assigning weights to each kernel function, and forming a linear combination according to the weights to obtain a multi-kernel model. In a specific application, the size of the effect of different kernel functions can be adjusted through weight adjustment so that the multi-kernel method can achieve the overall optimal effect in different occasions.

A multi-kernel model is obtained by linearly combining a variety of different kernel functions, where the type and number of kernel functions include but are not limited to the above four, and can also be any of the above four, which are not specifically limited here .

Based on the above-mentioned embodiments, the multi-core model is used to map the two sets of attribute variables corresponding to the optimal correlation coefficient to subspaces to obtain network operation data feature vectors, specifically including:

In this embodiment, the two sets of attribute variables corresponding to the optimal correlation coefficient are mapped to the subspace by each kernel function in the multi-kernel model to obtain two sets of mapped feature vectors.

In this embodiment, after obtaining the eigenvectors corresponding to each kernel function, the distance of the corresponding eigenvectors of each kernel function is calculated.

Based on the above-mentioned embodiment, the correlation of network operation data feature vectors obtained according to the distance and the weight of the kernel function specifically includes:

The correlation values of each kernel function are summed to obtain the correlation of the network operation data feature vector.

The details are shown in the following formula:

in,

is the feature vector after multi-core model mapping

and

the distance between;

Cr(A _i , B _i ) is the magnitude of the calculated correlation, and W _d is the weight corresponding to each kernel function.

Based on the above-mentioned embodiments, the optimal correlation coefficient of the attribute variable obtained based on canonical correlation analysis specifically includes:

In this embodiment, two sets of attribute variables are linearly combined, and a pair of coefficient vectors of the linear combination with the largest correlation coefficient are used to represent the correlation between the two sets of attribute variables. Two sets of attribute variables are linearly combined to obtain two sets of typical variables, and the coefficients of the linear combination are coefficient vectors. Among them, the relationship between attribute variables, coefficient vectors and canonical variables is as follows:

U＝a ₁ X ₁ +a ₂ X ₂ +...+a _p X _p ＝aX (3)

V＝b ₁ Y ₁ +b ₂ Y ₂ +...+b _q Y _q ＝bY (4)

Among them, U and V are typical variables after linear combination; a ₁ , a ₂ ... a _p and b ₁ , b ₂ ... b _q are coefficient vectors; X=(X ₁ , X ₂ ,..., X _P ) and Y=(Y ₁ , Y ₂ , ..., Y _q ) are attribute variables.

The correlation coefficient is used to describe the correlation between two groups of typical variables, and the larger the correlation coefficient is, the greater the correlation between the two groups of typical variables is. Among them, the relationship between the correlation coefficient and the canonical variables is as follows:

Cov (U, V) = a ^T Cov (X, Y) b = a ^T Σ ₁₂ b (6)

Var(U)＝a ^T Cov(X)a＝a ^T Σ ₁₁ a＝1 (7)

Var(V)＝b ^T Cov(Y)b＝b ^T Σ ₂₂ b＝1 (8)

Among them, ρ is the correlation coefficient between U and V, Cov(U, V) is the covariance of U and V, Var(U) is the sample variance of U, and Var(V) is the sample variance of V.

In the actual application process, the coefficient vector can be adjusted to change the value of ρ, so that the typical variables have a greater correlation.

Based on the above-mentioned embodiments, the correlation analysis method for smart city network resources provided by the present application further includes the following steps:

Specifically, in this embodiment, multiple sets of attribute variables of the acquired smart city network are combined in pairs according to the arrangement and combination manner. Carry out canonical correlation analysis on each combination separately, obtain the correlation coefficient of each combination, and confirm the best correlation coefficient with the largest correlation coefficient. Correspondingly, the combination with the largest correlation coefficient is mapped and correlated with multi-kernel functions.

Referring to FIG. 2 , the correlation analysis for smart city network resources provided by the present application will be specifically described below in conjunction with specific examples.

Step 210: using four normalized attribute vectors X1, X2, X3 and X4 of the same dimension;

in:

X1=(0.0068, 0.3573, 0.8925, . . . , 0.0391),

X2=(0.9432, 0.0033, 0.3819, . . . , 0.8239),

X3=(0.1670, 0.0329, 0.9028, ..., 0.6193),

X4 = (0.2931, 0.8352, 0.0091, . . . , 0.4890).

Divide the four vectors X1, X2, X3 and X4 into (X1, X2), (X1, X3), (X1, X4), (X2, X3), (X2, X4), (X3, X4) six Set of vector pairs. And substituting them into the following formulas (10)-(14), the a, b coefficients and the optimal ρ value corresponding to each vector pair are respectively obtained.

Taking the derivative of formula (9), we get:

After setting the derivative to zero, the system of equations is obtained:

Σ ₁₂ b-λΣ ₁₁ a=0 (12)

Σ ₂₁ a - θ Σ ₂₂ b = 0 (13)

In the above formula, the first equation is multiplied by a ^T on the left, the second equation is multiplied by b ^T on the right, and then according to a ^T ∑ ₁₁ a=1, b ^T ∑ ₂₂ b=1, λ=θ=a ^T ∑ ₁₂ b, λ is Corr(U, V), only the maximum λ is required.

Finally it can be deduced that:

According to the formulas (10)-(14), the obtained a, b coefficients and the optimal ρ value are shown in the following table:

Table 1:

Step 220: select a variety of different kernel functions, and construct a linear combination relationship between them;

As shown in the following formula:

Specifically, a polynomial kernel function K1, an exponential kernel function K2, a Gaussian kernel function K3 and a linear kernel function K4 are used to construct a linear combination of them multi-kernel function K, which is the multi-kernel model in this example. According to formula (15), respectively take θ ₁ =e, θ ₂ =2, θ ₃ =2e,

Among them, θ ₁ , θ ₂ , θ ₃ and θ ₄ are the weights corresponding to each kernel function, and the expression of the multi-kernel model is obtained by substituting the following formula:

Step 230: Using the four kernel functions in step 220, map the attribute with the largest correlation coefficient in step 210 to a 10-dimensional subspace to obtain the corresponding network operating data feature vector;

According to the data in the above table, the attribute combination with the largest correlation coefficient is (X2, X4). Therefore, the combination (X2, X4) is selected for mapping to obtain the feature vector of network operation data.

The network operation data feature vectors corresponding to the four kernel functions of K1, K2, K3 and K4 are as follows:

Step 240: Calculate the correlation between two groups of attributes according to the feature vector of network operation data.

Specifically, the distance between the feature vectors of the network operation data obtained after the vector combination (X2, X4) is mapped is calculated, so as to obtain the correlation between the two attributes.

Calculated by formula (1),

Taking e as the unit distance, set W _d as W ₁ =e, W ₂ =2e, W ₃ =4e, W ₄ =0.5e;

Cr(X2, X4)=25.4985e is calculated by the formula (2), that is, the correlation between the two groups of attributes is 25.4985e.

The following describes the correlation analysis device for smart city network resources provided by this application. The correlation analysis device for smart city network resources described below and the correlation analysis method for smart city network resources described above can be referred to each other .

Referring to FIG. 3, the present application provides a correlation analysis device for smart city network resources, including:

The optimal correlation coefficient acquisition module 310 is used to obtain multiple sets of different attribute variables of the smart city network, and obtain the optimal correlation coefficient of the attribute variables based on canonical correlation analysis;

The feature vector acquisition module 320 is configured to map the attribute variable corresponding to the optimal correlation coefficient to a subspace based on a multi-kernel model to obtain a feature vector of network operation data; wherein the multi-kernel model is established based on a linear combination of various kernel functions of;

The correlation analysis module 330 is configured to calculate the distance of the feature vector of the network operation data based on the Euclidean distance metric, and obtain the correlation of the feature vector of the network operation data according to the weight of the linear combination of the distance and the kernel function.

The correlation analysis device for smart city network resources provided by the embodiment of the present application obtains the optimal correlation coefficient of the attribute variable of the smart city network through typical correlation analysis, and maps the attribute variable corresponding to the optimal correlation coefficient to subspace. The present application can process nonlinear network data by combining multiple kernel functions with canonical correlation analysis, and obtain more accurate correlation of network data.

FIG. 4 illustrates a schematic diagram of the physical structure of an electronic device. As shown in FIG. 4, the electronic device may include: a processor (processor) 410, a communication interface (Communications Interface) 420, a memory (memory) 430 and a communication bus 440, Wherein, the processor 410 , the communication interface 420 , and the memory 430 communicate with each other through the communication bus 440 . The processor 410 can call the logic instructions in the memory 430 to execute a correlation analysis method for smart city network resources, the method includes:

In addition, the above logic instructions in the memory 430 may be implemented in the form of software function units and be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including Several instructions are used to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disc, etc., which can store program codes. .

On the other hand, the present application also provides a computer program product, the computer program product includes a computer program, the computer program can be stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer can Execute the correlation analysis method for smart city network resources provided by the above methods, the method includes:

In yet another aspect, the present application also provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it is implemented to perform the correlation of smart city network resources provided by the above methods. Analytical methods, which include:

The present invention also provides a chip, the chip includes a processor and a communication interface, the communication interface is coupled to the processor, and the processor is used to run programs or instructions to realize the smart city-oriented A correlation analysis method for network resources, the method comprising:

The device for analyzing the correlation of smart city network resources in the embodiment of the present application may be an electronic device, or a component in the electronic device, such as an integrated circuit or a chip. The electronic device may be a terminal, or other devices other than the terminal. Exemplarily, the electronic device can be a mobile phone, a tablet computer, a notebook computer, a handheld computer, a vehicle electronic device, a mobile Internet device (Mobile Internet Device, MID), an augmented reality (augmented reality, AR)/virtual reality (virtual reality, VR) ) equipment, robots, wearable devices, ultra-mobile personal computer (ultra-mobile personal computer, UMPC), netbook or personal digital assistant (personal digital assistant, PDA), etc., can also serve as server, network attached storage (Network Attached Storage, NAS), personal computer (personal computer, PC), television (television, TV), teller machine, or self-service machine, etc., which are not specifically limited in this embodiment of the present application.

The device for analyzing the correlation of smart city network resources in the embodiment of the present application may be a device with an operating system. The operating system may be an Android (Android) operating system, an ios operating system, or other possible operating systems, which are not specifically limited in this embodiment of the present application.

The memory 430 can be used to store software programs as well as various data. The memory 430 may mainly include a first storage area for storing programs or instructions and a second storage area for storing data, wherein the first storage area may store an operating system, an application program or instructions required by at least one function (such as a sound playing function, image playback function, etc.), etc. Also, the memory 430 may include volatile memory or nonvolatile memory, or, the memory 430 may include both volatile and nonvolatile memory. Among them, the non-volatile memory can be read-only memory (Read-Only Memory, ROM), programmable read-only memory (Programmable ROM, PROM), erasable programmable read-only memory (Erasable PROM, EPROM), electrically programmable Erase Programmable Read-Only Memory (Electrically EPROM, EEPROM) or Flash. Volatile memory can be random access memory (Random Access Memory, RAM), static random access memory (Static RAM, SRAM), dynamic random access memory (Dynamic RAM, DRAM), synchronous dynamic random access memory (Synchronous DRAM, SDRAM), double data rate synchronous dynamic random access memory (Double Data Rate SDRAM, DDRSDRAM), enhanced synchronous dynamic random access memory (Enhanced SDRAM, ESDRAM), synchronous connection dynamic random access memory (Synch link DRAM , SLDRAM) and Direct Memory Bus Random Access Memory (Direct Rambus RAM, DRRAM). The memory 430 in the embodiment of the present application includes but is not limited to these and any other suitable types of memory.

The device embodiments described above are only illustrative, and 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 it can be distributed to multiple network elements. Part or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. It can be understood and implemented by those skilled in the art without any creative efforts.

Through the above description of the implementations, those skilled in the art can clearly understand that each implementation can be implemented by means of software plus a necessary general-purpose hardware platform, and of course also by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the prior art can be embodied in the form of software products, and the computer software products can be stored in computer-readable storage media, such as ROM/RAM, magnetic discs, optical discs, etc., including several instructions to make a computer device (which may be a personal computer, server, or network device, etc.) execute the methods described in various embodiments or some parts of the embodiments.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, rather than limiting them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: it can still Modifications are made to the technical solutions described in the foregoing embodiments, or equivalent replacements are made to some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

A correlation analysis method for smart city network resources, including:

Obtain multiple sets of different attribute variables of the smart city network, and obtain the optimal correlation coefficient of the attribute variables based on canonical correlation analysis;

Mapping the attribute variable corresponding to the optimal correlation coefficient to a subspace based on a multi-kernel model to obtain a feature vector of network operation data; wherein the multi-kernel model is established based on a linear combination of multiple kernel functions;

The distance of the feature vector of the network operation data is calculated based on the Euclidean distance measure, and the correlation of the feature vector of the network operation data is obtained according to the weight of the linear combination of the distance and the kernel function.
The correlation analysis method for smart city network resources according to claim 1, wherein the multi-kernel model is established based on a linear combination of multiple kernel functions, specifically comprising:

Obtaining multiple types of kernel functions, linearly combining the kernel functions according to different weights, and obtaining a multi-kernel model after weight accumulation and linear combination of various kernel functions;

Wherein, the kernel function includes at least one of a polynomial kernel function, an exponential kernel function, a Gaussian kernel function and a linear kernel function.
The correlation analysis method for smart city network resources according to claim 1, wherein the multi-core model is used to map two sets of attribute variables corresponding to the optimal correlation coefficient to subspaces to obtain network operation data feature vectors, specifically include:

Based on each kernel function in the multi-kernel model, two sets of feature vectors corresponding to the two sets of attribute variables corresponding to the optimal correlation coefficient are mapped to the subspace;

The calculation of the distance of the feature vector of the network operation data based on the Euclidean distance metric specifically includes:

The distance between two sets of feature vectors corresponding to each kernel function is calculated based on the Euclidean distance metric.
According to the described correlation analysis method facing intelligent city network resource of claim 1, wherein, described according to the weight of described distance and described kernel function, obtains the correlation of network operating data feature vector, specifically comprises:

According to the distance between the weight of each kernel function in the multi-kernel model and the two groups of eigenvectors corresponding to each kernel function, the correlation value of each kernel function is obtained;

The correlation values of each kernel function are summed to obtain the correlation of the feature vectors of the network operation data.
According to the correlation analysis method for smart city network resources according to claim 1, wherein the optimal correlation coefficient of the attribute variable is obtained based on the canonical correlation analysis, specifically comprising:

Based on the canonical correlation analysis, the coefficient vector of the linear combination of the attribute variables and the typical variables after the linear combination are obtained;

Obtaining the correlation coefficient of the typical variable, adjusting the coefficient vector to make the correlation coefficient an optimal value, and obtaining the optimal correlation coefficient.
The correlation analysis method for smart city network resources according to any one of claims 1-5, further comprising:

Combining multiple sets of attribute variables in pairs, and obtaining the correlation coefficient corresponding to each combination;

The combination with the largest correlation coefficient is confirmed as the two groups of attribute variables corresponding to the optimal correlation coefficient.
A correlation analysis device for smart city network resources, including:

The optimal correlation coefficient acquisition module is used to obtain multiple sets of different attribute variables of the smart city network, and obtain the optimal correlation coefficient of the attribute variables based on canonical correlation analysis;

The feature vector acquisition module is used to map the attribute variable corresponding to the optimal correlation coefficient to the subspace based on the multi-kernel model to obtain the feature vector of the network operation data; wherein the multi-kernel model is established based on a linear combination of various kernel functions ;

The correlation analysis module is used to calculate the distance of the feature vector of the network operation data based on the Euclidean distance measure, and obtain the correlation of the feature vector of the network operation data according to the weight of the linear combination of the distance and the kernel function.
According to the relativity analysis device for smart city network resources according to claim 7, wherein the feature vector acquisition module is specifically used for:

Obtaining multiple types of kernel functions, linearly combining the kernel functions according to different weights, and obtaining a multi-kernel model after weight accumulation and linear combination of various kernel functions;

Wherein, the kernel function includes at least one of a polynomial kernel function, an exponential kernel function, a Gaussian kernel function and a linear kernel function.
According to the relativity analysis device for smart city network resources according to claim 7, wherein the feature vector acquisition module is also used for:

Based on each kernel function in the multi-kernel model, two sets of feature vectors corresponding to the two sets of attribute variables corresponding to the optimal correlation coefficient are mapped to the subspace;

The correlation analysis module is specifically used for: calculating the distance between two groups of feature vectors corresponding to each kernel function based on the Euclidean distance measure.
According to the correlation analysis device for smart city network resources according to claim 7, wherein the correlation analysis module is also used for:

According to the distance between the weight of each kernel function in the multi-kernel model and the two groups of eigenvectors corresponding to each kernel function, the correlation value of each kernel function is obtained;

The correlation values of each kernel function are summed to obtain the correlation of the feature vectors of the network operation data.
The correlation analysis device for smart city network resources according to claim 7, wherein the optimal correlation coefficient acquisition module is specifically used for:

Based on the canonical correlation analysis, the coefficient vector of the linear combination of the attribute variables and the typical variables after the linear combination are obtained;

Obtaining the correlation coefficient of the typical variable, adjusting the coefficient vector to make the correlation coefficient an optimal value, and obtaining the optimal correlation coefficient.
According to any one of claims 7-11, the correlation analysis device for smart city network resources, further comprising the following modules:

Combination module: combine multiple sets of attribute variables in pairs, and obtain the correlation coefficient corresponding to each combination;

Confirmation module: confirm the combination with the largest correlation coefficient as the two groups of attribute variables corresponding to the optimal correlation coefficient.
An electronic device, comprising a memory, a processor, and a computer program stored on the memory and operable on the processor, when the processor executes the program, the computer program described in any one of claims 1 to 6 is realized. The steps of the correlation analysis method for smart city network resources are described.
A non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the smart city network resource-oriented correlation analysis method according to any one of claims 1 to 6 is implemented step.
A computer program product, including a computer program, when the computer program is executed by a processor, the steps of the smart city network resource-oriented correlation analysis method according to any one of claims 1 to 6 are realized.
A chip, the chip includes a processor and a communication interface, the communication interface is coupled to the processor, the processor is used to run programs or instructions, and realize the intelligent city-oriented The steps of the correlation analysis method of network resources.