CN116362577A - Target class membership analysis method, system, device and storage medium - Google Patents

Abstract

The invention relates to a target class membership analysis method, a system, equipment and a storage medium, which comprise the following steps: analyzing different application scene characteristics and evaluation indexes of the item to be evaluated to obtain an evaluation index set and a scene characteristic set of the item to be evaluated, and respectively marking the evaluation index set and the scene characteristic set as a first variable set and a second variable set; performing membership degree analysis on the probability of each evaluation index in the first variable set and the probability of each scene characteristic in the second variable set under different application scenes to obtain an evaluation index list under different scene characteristics; and determining an evaluation index subset of the item to be evaluated under the preset scene characteristic based on the evaluation index list, and obtaining an evaluation result of the item to be evaluated under the preset scene characteristic based on the evaluation index subset. The invention quantitatively classifies the membership incidence relation of the occurrence of the fuzzy matters in the engineering through membership analysis based on the occurrence probability, and can be widely applied to the field of data mining classification of smart power grids.

Description

Method, system, device and storage medium for analyzing membership degree of target category

technical field

The invention relates to the field of data mining classification, in particular to the subordination relationship between factors in two types of related variable sets, and in particular to a method, system, device and storage medium for analyzing the membership degree of target categories based on variable association.

Background technique

At present, in the field of classification, methods such as cluster analysis, principal component analysis, expert scoring method, SVM analysis, etc., are all classified and analyzed from actual data. However, in the field of engineering analysis, there are many correlations that cannot be clearly defined. Events related to each other, such as the relationship between the frequency of transaction types and many factors, or the definition of the degree of membership between each evaluation index in a certain evaluation index system and the object category being evaluated, etc. The determination of such association relationship or fuzzy affiliation relationship is usually judged by methods such as expert scoring method, which is highly subjective.

Contents of the invention

In view of the problems referred to above, the object of the present invention is to provide a method, system, equipment and storage medium for analyzing the degree of membership of target categories based on variable association, through the analysis of the degree of membership based on the probability of occurrence, the degree of membership of the occurrence of fuzzy things occurring in engineering Quantitative classification of association relationship.

To achieve the above object, the present invention takes the following technical solutions:

In a first aspect, the present invention provides a target category membership analysis method, comprising the following steps:

Analyze the characteristics of different application scenarios and evaluation indicators of the items to be evaluated, obtain the evaluation index set and the scene characteristic set of the project to be evaluated, record the evaluation index set as the first variable set, and record the scene characteristic set as the second variable set;

The membership degree analysis is performed on the occurrence probability of each evaluation index in the first variable set and the probability of each scene characteristic in the second variable set under different application scenarios, and a list of evaluation indexes under different scene characteristics is obtained;

Based on the evaluation index list, the evaluation index subset of the item to be evaluated under the preset scene characteristics is determined, and the evaluation result of the item to be evaluated under the preset scene characteristics is obtained based on the evaluation index subset.

Further, performing membership degree analysis on the occurrence probability of each evaluation index in the first variable set under different application scenarios and the probability of each scene characteristic in the second variable set to obtain a list of evaluation indexes under different scene characteristics, including the following steps:

Constructing a variable probability matrix based on the probability of occurrence of each variable in the first variable set and the second variable set;

According to the variable probability matrix of the first variable set, based on the fuzzy analysis method, the ideal membership relationship of each evaluation index in the first variable set to the ideal scene characteristics is obtained; the ideal membership relationship means that each evaluation index in the first variable set is grouped in The probability of each ideal scene characteristic obtained according to the preset classification rules;

According to the variable probability matrix of the second variable set and the number of actual scene characteristics in the second variable set, the deviation correction is performed on the formed ideal membership relationship, and the relationship between each evaluation index in the first variable set and the actual scene characteristics in the second variable set is obtained. Actual affiliation, as a list of evaluation indicators under different scene characteristics.

Further, according to the variable probability matrix of the first variable set, the ideal membership relationship of each evaluation index in the first variable set to the ideal scene characteristics is obtained based on the fuzzy analysis method, including:

Determine the current number of cluster centers, and according to the determined current number of cluster centers, classify the variable probability matrix of the first variable set with the fuzzy K-means algorithm, and calculate the objective function value corresponding to the current number of cluster centers;

Calculate the cluster center matrix, membership matrix and objective function value corresponding to the number of all cluster centers;

The number of cluster centers with the smallest objective function value is selected as the number of ideal cluster centers, the corresponding cluster center matrix is the ideal cluster center matrix, and the corresponding membership degree matrix is used as the first variable set. ideal affiliation.

Further, the formula for calculating the objective function value is:

In the formula, J(U,W,c) is the objective function value calculated under the current membership degree matrix U, cluster center matrix W and the number of cluster centers c.

Further, according to the variable probability matrix of the second variable set and the number of actual scene characteristics in the second variable set, performing deviation correction on the formed ideal membership relationship, including:

Compare the variable probability matrix of the second variable set with the ideal cluster center matrix to obtain a distance matrix;

Using the fuzzy normalization method to normalize the obtained distance matrix to obtain the fuzzy membership degree matrix;

According to the number of actual scene characteristics in the second variable set and the comparison result of the number of ideal cluster centers, combined with the fuzzy membership degree matrix, the classification relationship is determined;

The evaluation indicators in the first variable set belonging to each ideal cluster center are assigned to the corresponding scene characteristics in the second variable set according to the obtained classification relationship, and the relationship between the evaluation indicators in the first variable set and the second variable set is obtained. The actual affiliation of the actual scene features in .

Further, the calculation formula of each element in the distance matrix is as follows:

d _kl ＝|w _k -w' _l |

In the formula, w _k is the probability value of occurrence of each actual scene characteristic in the second variable set, and w' _l is the element value in the ideal cluster center matrix.

Further, according to the comparison result of the number of actual scene characteristics in the second variable set and the number of ideal cluster centers, combined with the fuzzy membership matrix, the classification relationship is determined, including:

Compare the number q of the actual scene characteristics in the second variable set with the size of the ideal cluster center number C:

If q>C, for each scene characteristic in the second variable set, select the ideal cluster center of d' _kl =0 to classify;

If q=C, each scene feature in the second variable set corresponds to the ideal cluster center one by one;

If q<C, then for each ideal clustering center, select each scene characteristic in the second variable set with d _l ' _k =0 for classification.

In a second aspect, the present invention provides a target category membership analysis system, including:

The variable set determination module is used to analyze the characteristics of different application scenarios and evaluation indicators of the items to be evaluated, and obtain the evaluation index sets and scene characteristic sets of the items to be evaluated, which are respectively recorded as the first variable set and the second variable set;

The membership relationship determination module is used to analyze the membership degree of the occurrence probability of each evaluation index in the first variable set and the probability of each scene characteristic in the second variable set under different application scenarios, and obtain a list of evaluation indicators under different scene characteristics;

The evaluation module is used to determine the evaluation index subset of the project to be evaluated under the preset scene characteristics based on the obtained evaluation index list under different scene characteristics, and obtain the evaluation of the project to be evaluated under the preset scene characteristics based on the evaluation index subset result.

In a third aspect, the present invention provides a processing device, the processing device includes at least a processor and a memory, and a computer program is stored on the memory, and when the processor runs the computer program, it is executed to achieve the target category membership steps in the analytical method.

In a fourth aspect, the present invention provides a computer storage medium, on which computer readable instructions are stored, and the computer readable instructions can be executed by a processor to implement the steps of the target category membership analysis method.

Due to the adoption of the above technical solutions, the present invention has the following advantages: the present invention probabilizes the occurrence frequency of two types of affairs, then performs membership degree analysis based on the occurrence probability, and performs category optimization and deviation correction to realize the occurrence of fuzzy affairs Affiliation relationship, an innovative quantitative classification method for such problems that appear in large numbers in engineering at present.

Description of drawings

Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiment. The drawings are only for the purpose of illustrating a preferred embodiment and are not to be considered as limiting the invention. Throughout the drawings, the same reference numerals are used to refer to the same parts. In the attached picture:

FIG. 1 is a flow chart of a method for analyzing the membership degree of a target category based on variable association provided by an embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention more clear, the following will clearly and completely describe the technical solutions of the embodiments of the present invention in conjunction with the drawings of the embodiments of the present invention. Apparently, the described embodiments are some, not all, embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention belong to the protection scope of the present invention.

It should be noted that the terminology used here is only for describing specific implementations, and is not intended to limit the exemplary implementations according to the present application. As used herein, unless the context clearly dictates otherwise, the singular is intended to include the plural, and it should also be understood that when the terms "comprising" and/or "comprising" are used in this specification, they mean There are features, steps, operations, means, components and/or combinations thereof.

In some embodiments of the present invention, a method for analyzing the membership degree of target categories is provided. By constructing the probability correlation relationship between variables, using technologies such as fuzzy cluster analysis to judge the membership relationship between variable factors, and correcting the membership relationship category by the number of targets Finally, the quantification of the membership relationship between variables is carried out according to the membership degree value, which can provide reference and guidance for the correlation relationship between the components of any two types of events that may occur at the same time and the quantitative measurement of the relationship, and has a certain application prospect. .

Correspondingly, in some other embodiments of the present invention, a target category membership analysis system, device and medium are provided.

Example 1

As shown in Figure 1, a kind of target category membership analysis method provided by the present embodiment comprises the following steps:

S1. Analyze the characteristics of different application scenarios and evaluation indicators of the items to be evaluated, and obtain the evaluation index sets and scene characteristic sets of the items to be evaluated, which are respectively recorded as the first variable set and the second variable set;

S2. Perform membership degree analysis on the occurrence probability of each evaluation index in the first variable set under different application scenarios and the probability of each scene characteristic in the variable set B, and obtain a list of evaluation indexes under different scene characteristics;

S3. Based on the obtained list of evaluation indicators under different scene characteristics, determine the evaluation index subset of the item to be evaluated under the preset scene characteristics, and obtain the evaluation result of the item to be evaluated under the preset scene characteristics based on the evaluation index subset.

Preferably, in the above step S2, the membership degree analysis is performed on the occurrence probability of each evaluation index in the first variable set under different application scenarios and the probability of each scene characteristic in the second variable set to obtain a list of evaluation indexes under different scene characteristics, Include the following steps:

S11. Construct a variable probability matrix based on the occurrence probability of each variable in the first variable set and the second variable set;

S12. According to the variable probability matrix of the first variable set, based on the fuzzy analysis method, the ideal subordination relationship of each evaluation index in the first variable set to the ideal scene characteristics is obtained; wherein, the ideal subordination relationship means that each evaluation index in the first variable set is assigned to Set the probability of each ideal scene characteristic obtained according to the preset classification rules;

S13. According to the variable probability matrix of the second variable set and the quantity of the actual scene characteristics in the second variable set, the deviation correction is performed on the ideal membership relationship formed in step S12, and the comparison of each evaluation index in the first variable set to the second variable set is obtained. The actual affiliation relationship of the actual scene characteristics in , that is, the list of evaluation indicators under different scene characteristics.

Preferably, in the above step S11, the variable probability matrices of the first variable set and the second variable set are respectively calculated by comparing the variable factors in the first variable set, that is, the number of occurrences of evaluation indicators, and the variable factors in the second variable set, that is, each The number of occurrences of scene characteristics is counted to obtain the occurrence probability of each evaluation index in the first variable set and each scene characteristic in the second variable set, thereby obtaining two variable probability matrices.

Preferably, in the above step S12, the fuzzy analysis method may be a variety of classification techniques, including but not limited to the improved fuzzy K-means algorithm. Using the improved fuzzy K-means algorithm to obtain the ideal membership relationship of each evaluation index in the first variable set to the ideal scene characteristics, including:

S121. Determine the current number of cluster centers. In this embodiment, the number of cluster centers is determined by an enumeration method, enumerating from 2 to a preset value.

S122. According to the determined current number of cluster centers, classify the variable probability matrix of the first variable set with the fuzzy K-means algorithm, and calculate the objective function value corresponding to the current number of cluster centers.

Among them, the calculation formula of the objective function value is:

In the formula, J(U,W,c) is the objective function value calculated under the current membership degree matrix U, cluster center matrix W and the number of cluster centers c. Among them, the calculation formula of each matrix is:

V=[v ₁ ,...v _i ,...,v _n ] ^T ,i=1,2,...n (2)

W=[w ₁ ,...w _j ,...,w _c ] ^T ,j=1,2,...c (3)

为第L轮计算中隶属度矩阵中的元素，即第i个评价指标对第j个聚类中心的隶属度；W为聚类中心矩阵，w_j为聚类中心矩阵的元素；c为聚类中心数；V为变量集合A的变量概率矩阵，v_i表示第l轮计算中变量概率矩阵的元素v_i对聚类中心矩阵W中元素w_j的隶属度值，n为评价指标个数。In the formula, U ^L is the membership degree matrix in the fuzzy K-means algorithm,

is the element in the membership degree matrix in the L-th round of calculation, that is, the membership degree of the i-th evaluation index to the j-th clustering center; W is the clustering center matrix, w _j is the element of the clustering center matrix; c is the clustering center The number of cluster centers; V is the variable probability matrix of the variable set A, v _i represents the membership degree value of the element v _i of the variable probability matrix in the first round of calculation to the element w _j in the cluster center matrix W, and n is the number of evaluation indicators .

S123. Steps S121 to S122 are repeated to calculate the cluster center matrix, membership degree matrix and objective function value corresponding to all cluster center numbers.

S123. Select the number of cluster centers with the smallest objective function value as the number of ideal cluster centers C, the corresponding cluster center matrix is the ideal cluster center matrix, and the corresponding membership degree matrix is the pair of evaluation indicators in the first variable set Ideal affiliation for each ideal scene characteristic.

Preferably, in the above step S13, the method for correcting the deviation of the formed ideal membership relationship includes the following steps:

S131. Comparing the variable probability matrix of the second variable set with the ideal cluster center matrix to obtain a distance matrix.

Among them, comparing the variable probability matrix of the second variable set with the ideal cluster center matrix refers to comparing the value of each element in the variable probability matrix of the second variable set with the value of each cluster center element in the ideal cluster center matrix The distance is calculated, that is, the calculation formula of each element in the distance matrix is as follows:

d _kl ＝|w _k -w' _l | (5)

In the formula, w _k is the probability value of occurrence of each actual scene characteristic in the second variable set, and w' _l is the element value in the ideal cluster center matrix.

S132. Normalize the distance matrix obtained in step S131 by using a fuzzy normalization method to obtain a fuzzy membership degree matrix.

Among them, the calculation formula of each value in the fuzzy membership degree matrix is as follows:

In the formula, q is the number of variable factors in the second variable set, C is the number of ideal cluster centers, d' _kl and d' _lk are the fuzzy membership degrees of the l-th cluster center and the k-th scene characteristic, d _lk and d _kl are the distances between the l cluster center and the kth scene characteristic.

S133. According to the comparison result of the number of actual scene characteristics in the second variable set and the number of ideal cluster centers, combined with the fuzzy membership degree matrix, determine the classification relationship.

Specifically, compare the number q of the actual scene characteristics in the second variable set with the ideal cluster center number C:

If q>C, for each scene characteristic in the second variable set, select the ideal cluster center of d' _kl =0 to classify;

If q=C, each scene feature in the second variable set corresponds to the ideal cluster center one by one;

If q<C, then for each ideal clustering center, select each scene characteristic in the second variable set of d' _lk =0 to classify, if there is a certain clustering center corresponding to multiple scene characteristics (that is, there is k ₁ ≠ k ₂ , so that d' _lk1 =d' _lk2 =0), at this time, calculate the distance between the evaluation index of the first variable set belonging to the cluster center and each scene characteristic, and each index belongs to the nearest scene characteristic .

S134. Subordinate the evaluation indicators in the first variable set belonging to each ideal clustering center to the corresponding scene characteristics in the second variable set according to the classification relationship obtained in S133, and obtain the comparison of each evaluation index in the first variable set with respect to the second variable set. The actual membership relationship of the actual scene characteristics in the bivariate set, that is, the list of evaluation indicators under different scene characteristics.

Example 2

In this embodiment, the evaluation index of the digital system and its application scene target characteristics are taken as examples to introduce the target category membership analysis method proposed by the present invention in detail.

S1. Determine a set of evaluation indicators and a set of scene characteristics.

For example, for digital systems applied to digital infrastructure scenarios, the evaluation index set A is:

A＝[Work efficiency improvement degree, cost saving degree, platform reuse degree, data storage computing resource sharing degree, business processing speed, interface friendliness, ease of use, system response speed, lean management, operation and maintenance quality assurance level, ease of use Operability, business fit, application rate of each system, timeliness of system iteration, system reliability, data security, promotion value of digital model, storage resource utilization, corporate culture level, data quality]

For the digital infrastructure scenario, its target feature, that is, the scenario feature set B is:

B＝[Reliable operation support capability, basic support capability, strong applicability and good user experience]

S2. Determine a list of evaluation indicators under different scene characteristics.

S21. The probabilities of each evaluation index applied to the system evaluation obtained through expert investigation are as follows:

Probability matrix A'=[0.78, 0.5357, 0.464, 0.5, 0.5714, 0.4286, 0.5, 0.2143, 0.3571, 0.2857, 0.1429, 0.0714, 0.5, 0.4643, 0.1429, 0.1786, 0.1429, 0.1786]

Probability matrix B'=[0.244,0.402,0.084,0.024] obtained through expert investigation

S22. According to the obtained probability matrix, the subordination relationship of each index in the evaluation index set to the ideal category is obtained based on the fuzzy analysis technique.

In this embodiment, an improved fuzzy K-means algorithm is used to determine the affiliation relationship. First, the enumeration method is used to enumerate the number of optimal cluster centers; secondly, after the number of cluster centers is determined, the fuzzy clustering method is used to carry out index Clustering, judging the distance from the target feature, calculating the membership degree relationship of the index, performing iterations, selecting the ideal cluster center, and dividing the membership degree according to the current cluster.

Specifically, the following steps are included:

The enumeration method is used to cluster n evaluation indicators to obtain classification results.

Use c=2:m enumeration-clustering algorithm to iterate, the number of cluster centers is c, perform c clustering on the n evaluation indicators, get c cluster centers, iterate with enumeration method, and use n The probability of the occurrence of the evaluation index is used as the location point to classify, and the classification results of m clusters are obtained.

Specific steps include:

a. Determine the initial number of cluster centers as c=2.

b. Set the initial membership degree matrix U(0).

c. Calculate the cluster center matrix W.

d. Iterate the initial membership degree matrix according to the obtained cluster center matrix W to obtain a new membership degree matrix U(L+1).

e. Judging whether the convergence formula is satisfied according to the judgment formula (13): if it converges, the clustering center is determined, the clustering is completed, the objective function value is calculated, and the current index membership matrix, the position of the clustering center, and the objective function value are saved, and then Return to step a, take c=c+1 to continue iteration; otherwise, return to step c.

f. Compare the objective function values under the number of cluster centers, take the cluster center number with the smallest objective function value as the ideal cluster center number, and the corresponding cluster center position is the ideal cluster center position and membership degree matrix.

The analysis shows that when the number of cluster centers is 5, the value of the objective function is the smallest, and the membership matrix and ideal cluster center position at this time are obtained.

center＝[0.1876 0.5068 0.7822 0.131 0.3117] ^T

S23. Perform deviation correction on the established membership relationship according to the number of variable factors in the target feature set, and obtain the actual membership relationship of the evaluation index variable factors to the target feature variable factors. The calculation results are shown in Table 1 and Table 2 below.

Table 1 The affiliation relationship between scene characteristics and cluster centers

Table 2 Fuzzy membership degree

Based on the above calculations, according to the feature center classification method based on deviation tolerance, the corresponding evaluation indicators in the ideal cluster center matrix are divided into four categories, and after processing, the indicators belonging to the first cluster and the fourth cluster are uniformly put into the actual first cluster. Among the four types of feature centers, the second cluster corresponds to the actual second type feature centers, the third cluster corresponds to the actual third type feature centers, and the fifth cluster corresponds to the actual first type feature centers. Finally, the index list distribution of the digital infrastructure scenario is obtained as follows:

From the obtained membership degree matrix, it can be judged that each evaluation index in the index set belongs to [3 2 2 2 2 2 2 2 13 3 3 4 4 2 2 4 1 4 1] class respectively.

Then construct the actual affiliation relationship between each evaluation index in variable set A and each target feature in variable set B, and generate a scene-based feature quantification and index extraction list as shown in Table 3 below.

Table 3 Scenario-based feature quantification and index extraction list

S3. Based on the obtained actual affiliation relationship, evaluate the item to be evaluated under the preset scene characteristics.

Example 3

Embodiment 1 above provides a method for analyzing the degree of membership of a target category. Correspondingly, this embodiment provides a system for analyzing the degree of membership of a target category. The system provided in this embodiment can implement the target category membership analysis method in Embodiment 1, and the system can be realized by software, hardware or a combination of software and hardware. For example, the system may include integrated or separate functional modules or functional units to execute corresponding steps in the methods of Embodiment 1. Since the system of this embodiment is basically similar to the method embodiment, the description process of this embodiment is relatively simple. For relevant information, please refer to the part of the description of Embodiment 1. The embodiment of the system provided by this embodiment is only illustrative .

The target category membership analysis system provided in this embodiment includes:

The variable set determination module is used to analyze the characteristics of different application scenarios and evaluation indicators of the items to be evaluated, and obtain the evaluation index sets and scene characteristic sets of the items to be evaluated, which are respectively recorded as the first variable set and the second variable set;

The membership relationship determination module is used to analyze the membership degree of the occurrence probability of each evaluation index in the first variable set and the probability of each scene characteristic in the second variable set under different application scenarios, and obtain a list of evaluation indicators under different scene characteristics;

The evaluation module is used to determine the evaluation index subset of the project to be evaluated under the preset scene characteristics based on the obtained evaluation index list under different scene characteristics, and obtain the evaluation of the project to be evaluated under the preset scene characteristics based on the evaluation index subset result.

Example 4

This embodiment provides a processing device corresponding to the target category membership analysis method provided in Embodiment 1. The processing device may be a processing device for a client, such as a mobile phone, a notebook computer, a tablet computer, a desktop computer, etc. To carry out the method of embodiment 1.

The processing device includes a processor, a memory, a communication interface and a bus, and the processor, the memory and the communication interface are connected through the bus to complete mutual communication. A computer program that can run on the processor is stored in the memory, and the processor executes the target category membership analysis method provided in Embodiment 1 when running the computer program.

In some embodiments, the memory may be a high-speed random access memory (RAM: Random Access Memory), and may also include a non-volatile memory, such as at least one disk memory.

In some other embodiments, the processor may be a central processing unit (CPU), a digital signal processor (DSP) and other types of general-purpose processors, which are not limited herein.

Example 5

The target category membership analysis method in Embodiment 1 can be embodied as a computer program product, and the computer program product can include a computer-readable storage medium loaded with a method for performing the target category membership analysis described in Embodiment 1. computer readable program instructions for the method.

A computer readable storage medium may be a tangible device that holds and stores instructions for use by an instruction execution device. A computer readable storage medium may be, for example, but is not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination of the above.

Those skilled in the art should understand that the embodiments of the present invention may be provided as methods, systems, or computer program products. Accordingly, the present invention can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow chart or blocks of the flowchart and/or the block or blocks of the block diagrams. Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: it can still be 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 invention.

Claims (10)

1. The object class membership analysis method is characterized by comprising the following steps of:

analyzing different application scene characteristics and evaluation indexes of the item to be evaluated to obtain an evaluation index set and a scene characteristic set of the item to be evaluated, marking the evaluation index set as a first variable set, and marking the scene characteristic set as a second variable set;

performing membership degree analysis on the probability of each evaluation index in the first variable set and the probability of each scene characteristic in the second variable set under different application scenes to obtain an evaluation index list under different scene characteristics;

and determining an evaluation index subset of the item to be evaluated under the preset scene characteristic based on the evaluation index list, and obtaining an evaluation result of the item to be evaluated under the preset scene characteristic based on the evaluation index subset.

2. The method for analyzing membership of a target class according to claim 1, wherein the membership analysis is performed on the probability of occurrence of each evaluation index in the first variable set and the probability of occurrence of each scene characteristic in the second variable set under different application scenes to obtain an evaluation index list under different scene characteristics, and the method comprises the following steps:

constructing a variable probability matrix based on the probability of each variable in the first variable set and the second variable set;

obtaining ideal membership of each evaluation index in the first variable set to ideal scene characteristics based on a fuzzy analysis method according to the variable probability matrix of the first variable set; the ideal membership refers to the probability that all evaluation indexes in the first variable set are clustered in all ideal scene characteristics obtained according to a preset classification rule;

and carrying out deviation correction on the formed ideal membership according to the variable probability matrix of the second variable set and the number of the actual scene characteristics in the second variable set to obtain the actual membership of each evaluation index in the first variable set to the actual scene characteristics in the second variable set, and taking the actual membership as an evaluation index list under different scene characteristics.

3. The method of claim 2, wherein the obtaining, based on the fuzzy analysis method, the ideal membership of each evaluation index in the first variable set to the ideal scene characteristic according to the variable probability matrix of the first variable set comprises:

determining the number of current clustering centers, classifying the variable probability matrix of the first variable set according to the determined number of the current clustering centers by using a fuzzy K-means algorithm, and calculating to obtain a target function value corresponding to the number of the current clustering centers;

calculating to obtain a cluster center matrix, a membership matrix and a target function value corresponding to all the cluster center numbers;

and selecting the cluster center number with the smallest objective function value as an ideal cluster center number, wherein the corresponding cluster center matrix is an ideal cluster center matrix, and the corresponding membership matrix is used as an ideal membership of each evaluation index in the first variable set to each ideal scene characteristic.

4. The method of claim 3, wherein the objective function value calculation formula is:

wherein J (U, W, c) is the objective function value calculated under the conditions of the current membership degree matrix U, the clustering center matrix W and the clustering center number c.

5. The method for analyzing membership of a target class according to claim 3, wherein performing bias correction on the formed ideal membership according to the variable probability matrix of the second variable set and the number of actual scene characteristics in the second variable set includes:

comparing the variable probability matrix of the second variable set with the ideal clustering center matrix to obtain a distance matrix;

normalizing the obtained distance matrix by a fuzzy normalization method to obtain a fuzzy membership matrix;

determining a classification relation according to the number of actual scene characteristics in the second variable set and the comparison result of the ideal clustering center number by combining the fuzzy membership matrix;

and (3) membership of the evaluation indexes in the first variable set belonging to each ideal cluster center to corresponding scene characteristics in the second variable set according to the obtained classification relation, and obtaining the actual membership of each evaluation index in the first variable set to the actual scene characteristics in the second variable set.

6. The method of claim 5, wherein the formula for calculating each element in the distance matrix is as follows:

d _kl ＝|w _k -w' _l |

wherein w is _k For probability value, w 'of each actual scene characteristic occurrence in the second variable set' _l Is the element value in the ideal cluster center matrix.

7. The method of claim 5, wherein determining the classification relationship according to the comparison result of the number of actual scene characteristics and the number of ideal cluster centers in the second variable set, in combination with the fuzzy membership matrix, includes:

comparing the quantity q of the actual scene characteristics in the second variable set with the magnitude of the ideal cluster center number C:

if q>C, for each scene characteristic in the second set of variables, select d' _kl Ideal cluster center for =0;

if q=c, each scene characteristic in the second variable set corresponds to an ideal cluster center one by one;

if q<C, selecting d 'for each ideal cluster center' _lk Each scene characteristic in the second set of variables=0 is categorized.

8. A target class membership analysis system, comprising:

the variable set determining module is used for analyzing different application scene characteristics and evaluation indexes of the item to be evaluated to obtain an evaluation index set and a scene characteristic set of the item to be evaluated, which are respectively recorded as a first variable set and a second variable set;

the membership determining module is used for carrying out membership analysis on the probability of each evaluation index in the first variable set and the probability of each scene characteristic in the second variable set under different application scenes to obtain an evaluation index list under different scene characteristics;

the evaluation module is used for determining an evaluation index subset of the item to be evaluated under the preset scene characteristics based on the obtained evaluation index list under the different scene characteristics, and obtaining an evaluation result of the item to be evaluated under the preset scene characteristics based on the evaluation index subset.

9. A processing device comprising at least a processor and a memory, said memory having stored thereon a computer program, characterized in that the processor executes the steps of the object class membership analysis method according to any one of claims 1 to 7 when running said computer program.

10. A computer storage medium having stored thereon computer readable instructions executable by a processor to perform the steps of the target class membership analysis method according to any one of claims 1 to 7.

Images