Evaluation index data processing and evaluation method, computer equipment, medium and product

Abstract

The invention provides an evaluation index data processing and evaluation method, computer equipment, a medium and a product. In a specific embodiment, the evaluation index data processing method comprises the following steps: obtaining evaluation index data of a plurality of to-be-evaluated targets; the number of clustering centers of a K-means clustering algorithm is determined according to the sum of squares of errors corresponding to different numbers of clustering centers, the average contour coefficient corresponding to the number of the partial clustering centers and the number of preset evaluation grade values, and the number of the clustering centers of the K-means clustering algorithm is smaller than or equal to the number of the preset evaluation grade values; according to the number of clustering centers of the K-means clustering algorithm, grouping normalization based on the K-means clustering algorithm is carried out on the evaluation index data of the multiple to-be-evaluated targets, and evaluation grade values corresponding to the evaluation index data of the multiple to-be-evaluated targets are obtained. According to the embodiment, the objectivity of normalization of the evaluation index data and the rationality in physical significance can be improved, so that the reliability of an evaluation result is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, and more specifically, to an evaluation index data processing and evaluation method, computer equipment, medium and product. Background Art

[0002] In the scheme of evaluating objects such as equipment systems, it is usually necessary to combine multiple evaluation index data with different numerical ranges and dimensions for comprehensive evaluation. The evaluation result obtained by comprehensive evaluation is obtained by weighted aggregation of multiple evaluation index data. Since the evaluation index data contains multiple numerical ranges and dimensions, it is necessary to normalize the various evaluation index data before weighted aggregation.

[0003] Clustering is a common data analysis technique used to divide data into different groups so that the similarity between data in the same group is high, while the similarity between different groups is low. According to the basic idea of ​​clustering, the normalization of evaluation index data can be achieved through clustering algorithms. In related technologies, the schemes for normalizing various evaluation index data have the problems of introducing subjective errors, lack of objectivity, and lack of rationality in the physical sense. Summary of the invention

[0004] The purpose of the present disclosure is to provide an evaluation indicator data processing and evaluation method, computer equipment, medium and product to solve at least one of the problems existing in the prior art.

[0005] In order to achieve the above objectives, the present invention adopts the following technical solutions: The first aspect of the present disclosure provides an evaluation index data processing method, comprising: Obtain evaluation indicator data of multiple targets to be evaluated; Determine the number of cluster centers of the K-means clustering algorithm according to the sum of squares of errors corresponding to different numbers of cluster centers, the average silhouette coefficient corresponding to the number of partial cluster centers, and the number of preset evaluation grade values, wherein the number of cluster centers of the K-means clustering algorithm is less than or equal to the number of preset evaluation grade values; According to the number of cluster centers of the K-means clustering algorithm, the evaluation index data of the multiple targets to be evaluated are grouped and normalized based on the K-means clustering algorithm to obtain evaluation grade values ​​corresponding to the evaluation index data of the multiple targets to be evaluated.

[0006] Optionally, determining the number of cluster centers of the K-means clustering algorithm according to the sum of squared errors corresponding to different numbers of cluster centers, the average silhouette coefficient corresponding to the number of partial cluster centers, and the number of preset evaluation grade values ​​includes: Among the number of cluster centers greater than 3, the first N cluster centers whose change rates of the sum of squared errors corresponding to the number of cluster centers are arranged from large to small are used as candidate inflection point values; Among the N candidate inflection point values, select the candidate inflection point values ​​that meet conditions one and two as the number of candidate cluster centers, where: Condition 1: SSE(k n -1)-SSE(k n )<[SSE(2)-SSE(3)] / 2; Condition 2 is: k n ≤M; SSE(·) represents the sum of squared errors corresponding to the number of cluster centers·; k n represents the nth candidate inflection point value, 1≤n≤N; M is the number of evaluation level values, M>4; The average silhouette coefficient corresponding to each number of candidate cluster centers is calculated, and the number of candidate cluster centers with the largest average silhouette coefficient is determined as the number of cluster centers of the K-means clustering algorithm.

[0007] Optionally, the number N of the candidate inflection point values ​​is 2-4.

[0008] Optionally, the evaluation index data of the multiple targets to be evaluated are grouped and normalized based on the K-means clustering algorithm according to the number of cluster centers of the K-means clustering algorithm, and the evaluation grade values ​​corresponding to the evaluation index data of the multiple targets to be evaluated are obtained, including: According to the number of cluster centers of the K-means clustering algorithm, the evaluation index data of the multiple targets to be evaluated are clustered based on the K-means clustering algorithm, and the relationship between the number of cluster centers of the K-means clustering algorithm and the number of evaluation grade values ​​is determined: If the number of cluster centers of the K-means clustering algorithm is equal to the number of evaluation grade values, the evaluation index data consisting of the clustering results obtained by clustering based on the K-means clustering algorithm are grouped into groups corresponding to the evaluation grade values ​​corresponding to the evaluation index data of the multiple targets to be evaluated; If the number of cluster centers of the K-means clustering algorithm is less than the number of evaluation grade values, the cluster containing the most evaluation index data in the clustering result obtained by the K-means clustering algorithm is taken as the first cluster, and it is determined whether the corresponding average silhouette coefficient is monotonically decreasing when the number of cluster centers increases from 2 to the number of cluster centers of the K-means clustering algorithm: If not, the evaluation index data in the first cluster is divided into a plurality of first sub-clusters, and the evaluation index data consisting of the other clusters in the clustering result except the first cluster and the plurality of first sub-clusters are grouped to correspond to the evaluation grade values ​​corresponding to the evaluation index data of the plurality of targets to be evaluated, and the number of the first sub-clusters is set to make the number of groupings of the evaluation index data grouping equal to the number of the evaluation grade values; If so, the second cluster in the clustering results are merged into one cluster and the third cluster in the clustering results are merged into one cluster, the data value of the evaluation index data contained in the second cluster is smaller than the data value of the evaluation index data contained in the first cluster, and the data value of the evaluation index data contained in the third cluster is larger than the data value of the evaluation index data contained in the first cluster, the evaluation index data in the first cluster are divided into multiple first sub-clusters, and the evaluation index data composed of the merged second cluster, the merged third cluster and the multiple first sub-clusters are grouped to correspond to the evaluation level values ​​corresponding to the evaluation index data of the multiple targets to be evaluated, and the number of the first sub-clusters is set so that the number of groupings of the evaluation index data groupings is equal to the number of evaluation level values.

[0009] Optionally, the division scheme for dividing the evaluation index data in the first cluster into multiple first sub-clusters is a division scheme in which the data value of the evaluation index data contained in the j-th first sub-cluster is greater than the data value of the evaluation index data contained in the (j-1)-th first sub-cluster and the difference in the number of evaluation index data contained in each first sub-cluster is the smallest, j∈[2,J], J is the number of first sub-clusters.

[0010] A second aspect of the present disclosure provides an evaluation method, comprising: Obtain multiple evaluation indicator data for multiple targets to be evaluated; For each of the multiple evaluation indicator data of the multiple targets to be evaluated, respectively execute the evaluation indicator data processing method provided by the first aspect of the present disclosure to obtain an evaluation grade value corresponding to each of the multiple evaluation indicator data of the multiple targets to be evaluated; The evaluation grade value corresponding to each evaluation indicator data of an evaluation target is weighted and aggregated to obtain the evaluation grade value of the evaluation target.

[0011] Optionally, for each of the multiple evaluation indicator data, the number of preset evaluation level values ​​is the same.

[0012] A third aspect of the present disclosure provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method provided in the first aspect or the second aspect of the present disclosure is implemented.

[0013] A fourth aspect of the present disclosure is a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method provided in the first aspect or the second aspect of the present disclosure.

[0014] A fifth aspect of the present disclosure provides a computer program product, comprising a computer program, which, when executed by a processor, implements the method provided in the first aspect or the second aspect of the present disclosure.

[0015] The beneficial effects of the present disclosure are as follows: The technical solution disclosed in the present invention can realize dynamic grouping and normalization of evaluation index data, is applicable to evaluation index data with various distribution characteristics, and can improve the objectivity and physical rationality of the normalization of evaluation index data, thereby ensuring the reliability of the evaluation results of comprehensive evaluation of the target. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The specific implementation methods of the present disclosure are further described in detail below with reference to the accompanying drawings.

[0017] Figure 1 A flow chart of the evaluation index data processing method provided by an embodiment of the present disclosure is shown.

[0018] Figure 2 A graph showing the sum of squared errors and average contour coefficient of the first time data of 31 operating modes as a function of k value is shown.

[0019] Figure 3 A graph showing the sum of squared errors and average contour coefficient of the second time data of 31 operating modes as a function of k value is shown.

[0020] Figure 4 A schematic diagram showing the structure of a computer system for implementing the apparatus provided by an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0021] In order to more clearly illustrate the present disclosure, the present disclosure is further described below in conjunction with the embodiments and drawings. Similar components in the drawings are represented by the same reference numerals. It should be understood by those skilled in the art that the content described below is illustrative rather than restrictive, and should not be used to limit the scope of protection of the present disclosure.

[0022] like Figure 1 As shown, an embodiment of the present disclosure provides an evaluation index data processing method, comprising the following steps: S1. Obtain evaluation indicator data of multiple targets to be evaluated.

[0023] In a specific example, the object of evaluation is an equipment system, the target to be evaluated is the operating mode of the equipment system, and the evaluation index data is time-related data, safety protection-related data, or performance-related data of the operating mode of the equipment system.

[0024] S2. Determine the number of cluster centers of the K-means clustering algorithm according to the sum of squared errors corresponding to different numbers of cluster centers, the average silhouette coefficient corresponding to some numbers of cluster centers, and a preset number of evaluation grade values, wherein the number of cluster centers of the K-means clustering algorithm is less than or equal to a preset number of evaluation grade values.

[0025] Therefore, this embodiment combines the idea of ​​the elbow method and the silhouette coefficient method for determining the number of cluster centers of the K-means clustering algorithm, and determines the number of cluster centers of the K-means clustering algorithm by comprehensively considering the sum of squared errors and the average silhouette coefficient corresponding to the number of cluster centers, as well as the number of preset evaluation grade values. This can eliminate subjective errors, making the determined number of cluster centers more reasonable, and taking into account both the sum of squared errors and objectivity of the K-means clustering algorithm.

[0026] The following is a brief introduction to concepts such as the K-means clustering algorithm, the elbow method, and the silhouette coefficient method.

[0027] The K-means clustering algorithm is a distance-based, unsupervised clustering algorithm that divides data into k clusters through iterative solutions, and clusters data with similar characteristics into the same cluster, so that the similarity of data within the same cluster is high, and the similarity between different clusters is low. It can be seen that the evaluation index data can be grouped and normalized based on the K-means clustering algorithm, and different types of evaluation index data with different numerical ranges and different dimensions can be grouped and normalized into dimensionless evaluation grade values, such as A, B, C, D, and E, so as to achieve a comprehensive evaluation of objects such as equipment systems by combining multiple evaluation index data with different numerical ranges and different dimensions.

[0028] The steps of K-means clustering algorithm include: Step a: Randomly select k data as the initial clustering centers.

[0029] Step b: Calculate the distance between each data point and each cluster center, and assign each data point to the cluster center closest to it.

[0030] Step c: Update the center of each cluster to the mean of the data contained in the cluster.

[0031] Step d: Repeat steps b and c until the cluster center (or the center or centroid of the cluster) no longer changes.

[0032] According to the steps of the above K-means clustering algorithm, the K-means clustering algorithm needs to first determine the number of cluster centers, referred to as k value.

[0033] The elbow method is a method for determining the number of cluster centers. It selects the best k value by calculating the sum of squared errors (SSE) corresponding to different k values. The sum of squared errors represents the sum of squared distances from each data point to the cluster center to which it belongs, and can characterize the cohesion of the cluster. The elbow method looks for the inflection point where the sum of squared errors significantly decreases with the increase of k value. The inflection point can be considered as the best k value. Before the inflection point, increasing the k value will significantly reduce the sum of squared errors. After the inflection point, the contribution of increasing the k value to reducing the sum of squared errors becomes less obvious, and the effect of increasing the k value on reducing the sum of squared errors becomes smaller.

[0034] The silhouette coefficient method is another method to determine the number of cluster centers. For clustering tasks, we hope that the clusters obtained are as close as possible within the cluster and as far apart as possible. The silhouette coefficient is an evaluation index that combines cohesion and separation. The silhouette coefficient method calculates the silhouette coefficient of each data. The calculation of the silhouette coefficient involves two indicators: the average distance a from the data to other data in the same cluster and the minimum average distance b from the data to all data in other clusters. The silhouette coefficient of the data is (ba) / max(a,b). After calculating the silhouette coefficients of all data, we take the average value to get the average silhouette coefficient. . Average silhouette coefficient The value range of is [-1,1]. The closer the data distance within the cluster is and the farther the data distance between clusters is, the greater the average silhouette coefficient is. The closer it is to 1, the better the clustering effect. The silhouette coefficient method obtains the optimal k value by calculating the distance between clusters and the distance within clusters.

[0035] In a possible implementation, step S2 further includes: Among the number of cluster centers greater than 3, the first N cluster centers whose change rates of the sum of squared errors corresponding to the number of cluster centers are arranged from large to small are used as candidate inflection point values; Among the N candidate inflection point values, select the candidate inflection point values ​​that meet conditions one and two as the number of candidate cluster centers, where: Condition 1: SSE(k n -1)-SSE(k n )<[SSE(2)-SSE(3)] / 2; Condition 2 is: k n ≤M; SSE(·) represents the sum of squared errors corresponding to the number of cluster centers·; k n represents the nth candidate inflection point value, 1≤n≤N; M is the number of evaluation level values, M>4; The average silhouette coefficient corresponding to each number of candidate cluster centers is calculated, and the number of candidate cluster centers with the largest average silhouette coefficient is determined as the number of cluster centers of the K-means clustering algorithm.

[0036] Therefore, by first screening according to the rate of change of the square of the error corresponding to the number of cluster centers, and then screening the number of cluster centers screened out once according to the rate of change of the number of cluster centers screened out once and the preset number of evaluation grade values, and then determining the number of cluster centers selected by the K-means clustering algorithm according to the average silhouette coefficient corresponding to the number of cluster centers screened out twice, subjective errors can be effectively eliminated, making the determined number of cluster centers more reasonable, and effectively taking into account the sum of squared errors and objectivity of the K-means clustering algorithm. Moreover, since only the average silhouette coefficient corresponding to the number of cluster centers screened out twice needs to be calculated, the overall calculation amount is small, and it is suitable for data sets of various sizes.

[0037] In a possible implementation, the number N of candidate inflection point values ​​is 2 to 4. For example, the number N of candidate inflection point values ​​is 3.

[0038] For example, A screening to select candidate inflection point values: For k values ​​k>3, select the first three cluster center numbers ranked from large to small by the rate of change of the sum of squared errors (△SSE(k-1) / △SSE(k)) as candidate inflection point values ​​k1, k2 and k3, where △SSE(k-1)= SSE(k-1)-SSE(k), △SSE(k)= SSE(k)-SSE(k+1), SSE(k) represents the sum of squared errors corresponding to the number of clusters k, SSE(k-1) represents the sum of squared errors corresponding to the number of clusters (k-1), and SSE(k+1) represents the sum of squared errors corresponding to the number of clusters (k+1).

[0039] Secondary screening to select the number of candidate cluster centers: determine whether the candidate inflection point values ​​k1, k2 and k3 meet conditions one and two respectively. Assuming that the candidate inflection point values ​​k1 and k2 meet conditions one and two respectively, k3>M and k3 does not meet condition two, then the number of candidate cluster centers selected is k1 and k2.

[0040] Determine the number of cluster centers for the K-means clustering algorithm: Calculate the average silhouette coefficient corresponding to the number of candidate cluster centers k1 The average silhouette coefficient corresponding to the number of candidate cluster centers k2 , assuming > , then the number of candidate cluster centers k2 is determined as the number of cluster centers of the K-means clustering algorithm.

[0041] It can be understood that SSE(·) represents the sum of squared errors corresponding to the number of cluster centers·, that is, in condition 1, SSE(k n -1) indicates that the number of cluster centers is (k n -1) corresponding to the error sum of squares, SSE(k n ) indicates that the number of cluster centers is the candidate inflection point value k n The corresponding sum of squared errors, SSE(2) and SSE(3) represent the sum of squared errors when the number of cluster centers is 2 and 3, respectively.

[0042] S3. According to the number of cluster centers of the K-means clustering algorithm, the evaluation index data of the multiple targets to be evaluated are grouped and normalized based on the K-means clustering algorithm to obtain evaluation grade values ​​corresponding to the evaluation index data of the multiple targets to be evaluated.

[0043] In a possible implementation, step S3 further includes: clustering the evaluation index data of the multiple targets to be evaluated based on the K-means clustering algorithm according to the number of cluster centers of the K-means clustering algorithm, and determining the relationship between the number of cluster centers of the K-means clustering algorithm and the number of evaluation grade values: If the number of cluster centers of the K-means clustering algorithm is equal to the number of evaluation grade values, the evaluation index data consisting of the clustering results obtained by clustering based on the K-means clustering algorithm are grouped into groups corresponding to the evaluation grade values ​​corresponding to the evaluation index data of the multiple targets to be evaluated; If the number of cluster centers of the K-means clustering algorithm is less than the number of evaluation grade values, the cluster containing the most evaluation index data in the clustering result obtained by the K-means clustering algorithm is taken as the first cluster, and it is determined whether the corresponding average silhouette coefficient is monotonically decreasing when the number of cluster centers increases from 2 to the number of cluster centers of the K-means clustering algorithm: If not, the evaluation index data in the first cluster is divided into a plurality of first sub-clusters, and the evaluation index data consisting of the other clusters in the clustering result except the first cluster and the plurality of first sub-clusters are grouped to correspond to the evaluation grade values ​​corresponding to the evaluation index data of the plurality of targets to be evaluated, and the number of the first sub-clusters is set to make the number of groupings of the evaluation index data grouping equal to the number of the evaluation grade values; If so, the second cluster in the clustering results are merged into one cluster and the third cluster in the clustering results are merged into one cluster, the data value of the evaluation index data contained in the second cluster is smaller than the data value of the evaluation index data contained in the first cluster, and the data value of the evaluation index data contained in the third cluster is larger than the data value of the evaluation index data contained in the first cluster, the evaluation index data in the first cluster are divided into multiple first sub-clusters, and the evaluation index data composed of the merged second cluster, the merged third cluster and the multiple first sub-clusters are grouped to correspond to the evaluation level values ​​corresponding to the evaluation index data of the multiple targets to be evaluated, and the number of the first sub-clusters is set so that the number of groupings of the evaluation index data groupings is equal to the number of evaluation level values.

[0044] Therefore, according to whether the number of cluster centers of the K-means clustering algorithm is equal to the number of evaluation grade values, and whether the corresponding average silhouette coefficient is monotonically decreasing when the number of cluster centers increases from 2 to the number of cluster centers of the K-means clustering algorithm, dynamic grouping normalization of the evaluation index data can be achieved. This is suitable for evaluation index data with various distribution characteristics, such as evaluation index data with distribution characteristics that are concentrated at one end and sparse at the other end. It can improve the objectivity of the normalization of the evaluation index data and the rationality in the physical sense, thereby ensuring the reliability of the evaluation results of comprehensive evaluation of the target.

[0045] Among them, judging whether the average silhouette coefficient corresponding to the increase in the number of cluster centers from 2 to the number of cluster centers of the K-means clustering algorithm is monotonically decreasing can be expressed as: judging the average silhouette coefficient (i)> (i+1), i∈[2,k x ) is established, where (i) is the average silhouette coefficient corresponding to the number of cluster centers i, (i+1) is the average silhouette coefficient corresponding to the number of cluster centers (i+1), k x The number of cluster centers of the K-means clustering algorithm determined in step S2.

[0046] For the above implementation of step S3, For example, if the number of evaluation grade values ​​is 5 and the number of cluster centers of the K-means clustering algorithm is also 5, then the 5 groups of evaluation index data consisting of the 5 clusters obtained by the clustering results based on the K-means clustering algorithm can be directly corresponded to 5 evaluation grade values. Each evaluation grade value group is normalized to the evaluation grade value corresponding to the group to which it belongs. The 5 evaluation grade values ​​can be expressed as A, B, C, D, E, etc.

[0047] For another example, the number of evaluation grade values ​​is 5, the number of cluster centers of the K-means clustering algorithm is 4, and the corresponding average silhouette coefficient is non-monotonically decreasing when the number of cluster centers increases from 2 to the number of cluster centers of the K-means clustering algorithm. Then: the clustering result includes 4 clusters, which are arranged from small to large according to the data values ​​of the evaluation index data contained therein, namely the 1st cluster, the 2nd cluster, the 3rd cluster and the 4th cluster. Assuming that the cluster containing the most evaluation index data is the 3rd cluster, that is, the 3rd cluster is taken as the first cluster, the evaluation index data in the 3rd cluster is divided into two first subclusters, respectively recorded as the 1st subcluster and the 2nd subcluster, and the 5 groups of evaluation index data consisting of the 1st cluster, the 2nd cluster, the 1st subcluster, the 2nd subcluster and the 4th cluster correspond to 5 evaluation grade values.

[0048] For another example, the number of evaluation grade values ​​is 5, the number of cluster centers of the K-means clustering algorithm is 4, and the corresponding average silhouette coefficient is monotonically decreasing when the number of cluster centers increases from 2 to the number of cluster centers of the K-means clustering algorithm. Then: the clustering result includes 4 clusters, which are arranged from small to large according to the data values ​​of the evaluation index data contained therein, namely the 1st cluster, the 2nd cluster, the 3rd cluster and the 4th cluster. Assuming that the cluster containing the most evaluation index data is the 3rd cluster, that is, the 3rd cluster is taken as the first cluster, the 1st cluster and the 2nd cluster are merged into a 1st merged cluster, and the evaluation index data in the 3rd cluster is divided into three first subclusters, which are respectively recorded as the 1st subcluster, the 2nd subcluster and the 3rd subcluster, and the 5 groups of evaluation index data consisting of the 1st merged cluster, the 1st subcluster, the 2nd subcluster, the 3rd subcluster and the 4th cluster correspond to 5 evaluation grade values.

[0049] For another example, the number of evaluation grade values ​​is 5, the number of cluster centers of the K-means clustering algorithm is 3, and the corresponding average silhouette coefficient is monotonically decreasing when the number of cluster centers increases from 2 to the number of cluster centers of the K-means clustering algorithm. Then: the clustering result includes 3 clusters, which are arranged from small to large according to the data values ​​of the evaluation index data contained therein, namely the 1st cluster, the 2nd cluster and the 3rd cluster. Assuming that the cluster containing the most evaluation index data is the 1st cluster, that is, the 1st cluster is taken as the first cluster, the 2nd cluster and the 3rd cluster are merged into a 1st merged cluster, and the evaluation index data in the 1st cluster is divided into four first subclusters, which are respectively recorded as the 1st subcluster, the 2nd subcluster, the 3rd subcluster and the 4th subcluster, and the 5 groups of evaluation index data consisting of the 1st subcluster, the 2nd subcluster, the 3rd subcluster, the 4th subcluster and the 1st merged cluster correspond to 5 evaluation grade values.

[0050] In a possible implementation, the division scheme for dividing the evaluation index data in the first cluster into multiple first sub-clusters is a division scheme in which the data value of the evaluation index data contained in the j-th first sub-cluster is greater than the data value of the evaluation index data contained in the (j-1)-th first sub-cluster and the difference in the number of evaluation index data contained in each first sub-cluster is the smallest, j∈[2,J], J is the number of first sub-clusters.

[0051] The division scheme of the present implementation method can also be expressed as evenly distributing the non-repetitive evaluation index data in the first cluster to multiple first sub-clusters. For example, the first cluster includes 15 evaluation index data, and the data is, for example, weight value, with the dimension of kilogram. The first cluster needs to be divided into two first sub-clusters, and the data value set of the first cluster is expressed as {1,1,1,1,2,2,2,2,2,3,3,3,4,4,4}. Then, according to the division scheme of the present implementation method, the first first sub-cluster obtained is {1,1,1,1,2,2,2,2,2}, and the second first sub-cluster is {3,3,3,4,4,4}.

[0052] It is understandable that the characteristics of the evaluation index data in a cluster are reflected in the data value. Therefore, when a cluster is divided into subclusters, it is necessary to divide them according to the size of the data value to comply with the clustering idea.

[0053] The following two specific examples illustrate steps S1 to S3 of the evaluation index data processing method provided in this embodiment.

[0054] In both specific examples, the evaluation object is the equipment system, the target to be evaluated is the 31 operating modes of the equipment system, and the preset number of evaluation level values ​​is 5.

[0055] , 108.44, 108.73, 112.30, 112.77, 112.91, 112.77, 112.97, 112.78, 112.79, 112.91, 112.77, 112.97, 112.77, 112.97, 112.77, 112.97, 112.77, 112.97, 112.77, 112.97, 112.77, 112.97, 112.77, 112.97, 112.77, 112.97, 112.77, 112.97, 112.77, 112.97, After calculating the sum of squared errors and average contour coefficients, the curves of the sum of squared errors and average contour coefficients of the first time data of 31 operating modes as a function of k value are shown in the figure below: Figure 2 As shown, Figure 2In the figure, the curve of the circular data point is the curve of the error sum of squares changing with the k value, and the curve of the star-shaped data point is the curve of the average silhouette coefficient changing with the k value. Based on the judgment and screening of step S2, the number of cluster centers of the K-means clustering algorithm is determined to be 5. Since the number of cluster centers of the K-means clustering algorithm is 5, the number of evaluation grade values ​​is 5. Therefore, the 5 groups of evaluation index data consisting of the 5 clusters included in the clustering result obtained by the clustering based on the K-means clustering algorithm can be directly corresponded to 5 evaluation grade values, which are: Group 1 {31.55, 39.12, 42.19, 43.68}, the corresponding evaluation grade value of Group 1 is 0.2; Group 2 {62.95, 69.52, 71.09, 76.68, 76.68, 84.01, 88.79, 94.79, 98.51, 102.12}, the corresponding evaluation grade value of Group 2 is 0.4; Group 3 {119.07, 119.27, 121.01, 129.00, 129.00, 138.16, 160.03}, the corresponding evaluation grade value of Group 3 is 0.6; Group 4 {191.46, 191.46, 195.30, 195.64, 200.09, 220.62}, the corresponding evaluation grade value of Group 4 is 0.8; Group 5 {356.90, 356.90, 410.90, 428.25}, the corresponding evaluation level value of Group 5 is 1.

[0056] In the second specific example, the evaluation index data obtained are 31 second time data of 31 operating modes of the equipment system, with the dimension of seconds, and the data set is {45.92, 45.92, 45.92, 45.92, 326.00, 326.00, 349.73, 349.73, 349.73, 379.71, 385.68, 385.68 , 385.68, 385.68, 428.07, 428.07, 428.07, 428.07, 428.07, 494.28, 499.74, 513.13, 520.13, 520.63, 554.19, 1030.42, 1030.42, 1030.42, 1030.42, 1030.42}. After calculating the sum of squares of the error and the average contour coefficient, the curves of the sum of squares of the error and the average contour coefficient of the second time data of 31 operating modes changing with the k value are shown in the figure below. Figure 3 As shown, Figure 3In the figure, the curve of the circular data point is the curve of the error sum of squares changing with the k value, and the curve of the star-shaped data point is the curve of the average silhouette coefficient changing with the k value. Based on the judgment and screening in step S2, the number of cluster centers of the K-means clustering algorithm is determined to be 4, and the number of cluster centers of the K-means clustering algorithm 4 is less than the number of evaluation level values ​​5. The clustering result obtained by clustering based on the K-means clustering algorithm includes a set of data values ​​of 4 clusters as shown below: Cluster 1: {45.92, 45.92, 45.92, 45.92}; Cluster 2 {326.00, 326.00, 349.73, 349.73, 349.73, 379.71, 385.68, 385.68, 385.68, 385.68, 428.07, 428.07, 428.07, 428.07}; The third cluster {494.28, 499.74, 513.13, 520.13, 520.63, 554.19}; The 4th cluster {1030.42, 1030.42, 1030.42, 1030.42, 1030.42}.

[0057] Among the four clusters, the second cluster contains the largest number of evaluation index data. (i)> (i+1), i∈[2,3) does not hold (e.g. Figure 3 As shown, (2)< (3)), so the first cluster, the third cluster and the fourth cluster remain unchanged, and the second cluster is divided into the first sub-cluster and the second sub-cluster, which are: The first subcluster is {326.00, 326.00, 349.73, 349.73, 349.73, 379.71}; The second subcluster is {385.68, 385.68, 385.68, 385.68, 428.07, 428.07, 428.07, 428.07, 428.07}.

[0058] The five groups of evaluation index data consisting of the first cluster, the first sub-cluster, the second sub-cluster, the third cluster, and the fourth cluster correspond to five evaluation level values, which are: Group 1 {45.92, 45.92, 45.92, 45.92}, the corresponding evaluation grade value of Group 1 is 0.2; Group 2 {326.00, 326.00, 349.73, 349.73, 349.73, 379.71}, the corresponding evaluation grade value of Group 2 is 0.4; Group 3 {385.68, 385.68, 385.68, 385.68, 428.07, 428.07, 428.07, 428.07, 428.07}, the corresponding evaluation grade value of Group 3 is 0.6; Group 4 {494.28, 499.74, 513.13, 520.13, 520.63, 554.19}, the corresponding evaluation grade value of Group 4 is 0.8; The 5th group is {1030.42, 1030.42, 1030.42, 1030.42, 1030.42}, and the corresponding evaluation level value of the 5th group is 1.

[0059] Another embodiment of the present disclosure provides an evaluation method, comprising the following steps: Obtain multiple evaluation indicator data for multiple targets to be evaluated; For each of the multiple evaluation indicator data of the multiple targets to be evaluated, the evaluation indicator data processing method provided in the above embodiment is respectively performed to obtain an evaluation grade value corresponding to each of the multiple evaluation indicator data of the multiple targets to be evaluated; The evaluation grade value corresponding to each evaluation indicator data of an evaluation target is weighted and aggregated to obtain the evaluation grade value of the evaluation target.

[0060] In a possible implementation, for each of the multiple evaluation index data, the number of preset evaluation level values ​​is the same, thereby improving the rationality of weighted aggregation and thus improving the reliability of the evaluation result.

[0061] In a specific example, continuing with the previous example, the object of evaluation is the equipment system, and the target to be evaluated is the 31 operating modes of the equipment system. If the operating mode is evaluated based on the first time data and the second time data, the evaluation level a corresponding to the first time data (88.79) of the first operating mode is 0.4, and the evaluation level b corresponding to the second time data (1030.42) is 1. Assuming that the weight α corresponding to the first time data is 0.3 and the weight β corresponding to the second time data is 0.7, the evaluation level value of the first operating mode is α×a+β×b=0.3×0.4+0.7×1=0.82.

[0062] The above example is an example of obtaining the evaluation grade value by directly weighting and aggregating the evaluation index data after grouping and normalizing. In addition, a multi-layer index system may also be adopted, where the evaluation index data is the bottom-layer index, and the evaluation index data as the bottom-layer index is grouped and normalized and then weighted and aggregated to obtain the second-layer index, and then weighted and aggregated layer by layer to obtain the final evaluation grade value, which is not limited in this embodiment.

[0063] like Figure 4 As shown, a computer system suitable for implementing the evaluation index data processing method provided by the above embodiment or the evaluation method provided by the above embodiment includes a central processing module (CPU), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) or a program loaded from a storage part into a random access memory (RAM). Various programs and data required for the operation of the computer system are also stored in the RAM. The CPU, ROM and RAM are connected to each other through a bus. An input / output (I / O) interface is also connected to the bus.

[0064] The following components are connected to the I / O interface: an input section including a keyboard, a mouse, etc.; an output section including a liquid crystal display (LCD), etc. and a speaker, etc.; a storage section including a hard disk, etc.; and a communication section including a network interface card such as a LAN card, a modem, etc. The communication section performs communication processing via a network such as the Internet. A drive is also connected to the I / O interface as needed. Removable media, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., are installed on the drive as needed so that a computer program read therefrom is installed into the storage section as needed.

[0065] In particular, according to this embodiment, the process described in the flowchart above can be implemented as a computer software program. For example, this embodiment includes a computer program product, which includes a computer program tangibly contained on a computer-readable medium, and the computer program includes program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication part, and / or installed from a removable medium.

[0066] The flow chart and schematic diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product of the present embodiment. In this regard, each square box in the flow chart or schematic diagram can represent a part of a module, program segment or code, and a part of the above-mentioned module, program segment or code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the schematic diagram and / or flow chart, and the combination of the square boxes in the schematic diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0067] As another aspect, this embodiment further provides a non-volatile computer storage medium, which may be the non-volatile computer storage medium included in the above-mentioned device in the above-mentioned embodiment, or may be a non-volatile computer storage medium that exists independently and is not installed in the terminal. The above-mentioned non-volatile computer storage medium stores one or more programs, and when the above-mentioned one or more programs are executed by a device, the above-mentioned device implements the evaluation index data processing method provided in the above-mentioned embodiment or the evaluation method provided in the above-mentioned embodiment.

[0068] As another aspect, the present embodiment further provides a computer program product, including a computer program, which, when executed by a processor, implements the evaluation indicator data processing method provided in the above embodiment or the evaluation method provided in the above embodiment.

[0069] In the description of the present disclosure, it should be noted that the orientation or positional relationship indicated by the terms "upper", "lower", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present disclosure and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present disclosure. Unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, and it can be the internal connection of two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present disclosure can be understood according to the specific circumstances.

[0070] It should also be noted that, in the description of the present disclosure, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or equipment. In the absence of further restrictions, the elements defined by the statement "comprise one..." do not exclude the presence of other identical elements in the process, method, article or equipment including the elements.

[0071] Obviously, the above embodiments of the present disclosure are merely examples for clearly illustrating the present disclosure, and are not limitations on the implementation methods of the present disclosure. For ordinary technicians in this field, other different forms of changes or modifications can be made on the basis of the above description. It is impossible to list all the implementation methods here. All obvious changes or modifications derived from the technical solution of the present disclosure are still within the protection scope of the present disclosure.

Claims

1. A method for processing evaluation index data, characterized in that: include: Obtain evaluation indicator data of multiple targets to be evaluated; Determine the number of cluster centers of the K-means clustering algorithm according to the sum of squares of errors corresponding to different numbers of cluster centers, the average silhouette coefficient corresponding to the number of partial cluster centers, and the number of preset evaluation grade values, wherein the number of cluster centers of the K-means clustering algorithm is less than or equal to the number of preset evaluation grade values; According to the number of cluster centers of the K-means clustering algorithm, the evaluation index data of the multiple targets to be evaluated are grouped and normalized based on the K-means clustering algorithm to obtain evaluation grade values ​​corresponding to the evaluation index data of the multiple targets to be evaluated.

2. The method according to claim 1, characterized in that The method of determining the number of cluster centers of the K-means clustering algorithm according to the sum of squared errors corresponding to different numbers of cluster centers, the average silhouette coefficient corresponding to the number of partial cluster centers, and the number of preset evaluation grade values ​​includes: Among the number of cluster centers greater than 3, the first N cluster centers whose change rates of the sum of squared errors corresponding to the number of cluster centers are arranged from large to small are used as candidate inflection point values; Among the N candidate inflection point values, select the candidate inflection point values ​​that meet conditions 1 and 2 as the number of candidate cluster centers, where: Condition 1: SSE(k n -1)-SSE(k n )<[SSE(2)-SSE(3)] / 2; Condition 2 is: k n ≤M; SSE(·) represents the sum of squared errors corresponding to the number of cluster centers·; k n represents the nth candidate inflection point value, 1≤n≤N; M is the number of evaluation level values, M>4; The average silhouette coefficient corresponding to each number of candidate cluster centers is calculated, and the number of candidate cluster centers with the largest average silhouette coefficient is determined as the number of cluster centers of the K-means clustering algorithm.

3. The method according to claim 2, characterized in that The number N of the candidate inflection point values ​​is 2-4.

4. The method according to any one of claims 1 to 3, characterized in that According to the number of cluster centers of the K-means clustering algorithm, the evaluation index data of the multiple targets to be evaluated are grouped and normalized based on the K-means clustering algorithm to obtain the evaluation grade values ​​corresponding to the evaluation index data of the multiple targets to be evaluated, including: According to the number of cluster centers of the K-means clustering algorithm, the evaluation index data of the multiple targets to be evaluated are clustered based on the K-means clustering algorithm, and the relationship between the number of cluster centers of the K-means clustering algorithm and the number of evaluation grade values ​​is determined: If the number of cluster centers of the K-means clustering algorithm is equal to the number of evaluation grade values, the evaluation index data consisting of the clustering results obtained by clustering based on the K-means clustering algorithm are grouped into groups corresponding to the evaluation grade values ​​corresponding to the evaluation index data of the multiple targets to be evaluated; If the number of cluster centers of the K-means clustering algorithm is less than the number of evaluation grade values, the cluster containing the most evaluation index data in the clustering result obtained by the K-means clustering algorithm is taken as the first cluster, and it is determined whether the corresponding average silhouette coefficient is monotonically decreasing when the number of cluster centers increases from 2 to the number of cluster centers of the K-means clustering algorithm: If not, the evaluation index data in the first cluster is divided into a plurality of first sub-clusters, and the evaluation index data consisting of the other clusters in the clustering result except the first cluster and the plurality of first sub-clusters are grouped to correspond to the evaluation grade values ​​corresponding to the evaluation index data of the plurality of targets to be evaluated, and the number of the first sub-clusters is set to make the number of groupings of the evaluation index data grouping equal to the number of the evaluation grade values; If so, the second cluster in the clustering results are merged into one cluster and the third cluster in the clustering results are merged into one cluster, the data value of the evaluation index data contained in the second cluster is smaller than the data value of the evaluation index data contained in the first cluster, and the data value of the evaluation index data contained in the third cluster is larger than the data value of the evaluation index data contained in the first cluster, the evaluation index data in the first cluster are divided into multiple first sub-clusters, and the evaluation index data composed of the merged second cluster, the merged third cluster and the multiple first sub-clusters are grouped to correspond to the evaluation level values ​​corresponding to the evaluation index data of the multiple targets to be evaluated, and the number of the first sub-clusters is set so that the number of groupings of the evaluation index data groupings is equal to the number of evaluation level values.

5. The method according to claim 4, characterized in that The division scheme for dividing the evaluation index data in the first cluster into multiple first sub-clusters is a division scheme in which the data value of the evaluation index data contained in the j-th first sub-cluster is greater than the data value of the evaluation index data contained in the (j-1)-th first sub-cluster and the difference in the number of evaluation index data contained in each first sub-cluster is the smallest, j∈[2,J], J is the number of first sub-clusters.

6. An evaluation method, characterized in that include: Obtain multiple evaluation indicator data for multiple targets to be evaluated; For each of the multiple evaluation indicator data of the multiple targets to be evaluated, respectively execute the evaluation indicator data processing method described in any one of claims 1 to 5 to obtain an evaluation grade value corresponding to each of the multiple evaluation indicator data of the multiple targets to be evaluated; The evaluation grade value corresponding to each evaluation indicator data of an evaluation target is weighted and aggregated to obtain the evaluation grade value of the evaluation target.

7. The method according to claim 6, characterized in that For each of the multiple evaluation index data, the number of preset evaluation level values ​​is the same.

8. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 5 is implemented, or the method according to claim 6 or 7 is implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 5 is implemented, or the method according to claim 6 or 7 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented, or the method according to claim 6 or 7 is implemented.

Images