Method and device for determining network evaluation index score and network input energy efficiency portrait

Abstract

The application relates to the technical field of data processing, and provides a network evaluation index scoring and network input energy efficiency portrait determination method and device. The method comprises the following steps: acquiring network evaluation index data of an evaluation object; determining a benchmark value of the network evaluation index data based on preset parameters; if the network evaluation index data is linear data, scoring the network evaluation index data according to a preset scoring model and the benchmark value to obtain a first data score value; if the network evaluation index data is nonlinear data, scoring the network evaluation index data based on a preset nonlinear scoring model and the benchmark value to obtain a second data score value; and obtaining a scoring result of the evaluation object according to the first data score value and the second data score value. The application can reduce the strong dependence of a scoring model on a maximum value or a minimum value, and can realize adjustment of a scoring result through preset parameters and scoring models of different data types, so that the evaluation processing requirements of various network evaluation indexes can be met.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method and apparatus for determining network evaluation index scoring and network input energy efficiency profile. Background Technology

[0002] In the process of creating network investment efficiency profiles for different cities, it is often necessary to evaluate and profile the network from multiple dimensions, including customers, business volume, revenue, and benefits. This requires both improving the comparability of evaluation indicators and ensuring the evaluation results have good discriminatory power. Processing network evaluation indicator data, especially the normalization and scoring of evaluation indicator data, has always been a technical challenge in the evaluation field.

[0003] To ensure the comparability of network health assessments or profiles, it is necessary to process the assessment indicators through scoring. Currently, the most common scoring method is linear scoring. Linear scoring uses the maximum and minimum values ​​in the dataset as scoring benchmarks, for example, a maximum score of 100 and a minimum score of 0. After linearly mapping the original data to a specified range, the assessment indicator score remains fixed and cannot be flexibly adjusted. The assessment results produced by this method are relatively singular and cannot meet the needs of assessing multiple types of network indicators. Summary of the Invention

[0004] This application provides a method and apparatus for determining network evaluation index scoring and network input energy efficiency profile, which solves the shortcomings of existing linear scoring methods that map raw data linearly to a specified range, resulting in fixed and unchanging evaluation index scores, inflexible evaluation index scores, and relatively singular evaluation results that cannot meet the evaluation processing needs of multiple types of network evaluation indicators.

[0005] In a first aspect, this application provides a method for scoring network evaluation metrics, comprising: acquiring network evaluation metric data of an evaluation object; determining a benchmark value of the network evaluation metric data based on preset parameters, wherein the preset parameters are determined based on the probability density of the network evaluation metric data; if the network evaluation metric data is linear data, scoring the network evaluation metric data according to a preset scoring model and the benchmark value to obtain a first data score value; if the network evaluation metric data is nonlinear data, scoring the network evaluation metric data based on a preset nonlinear scoring model and the benchmark value to obtain a second data score value; and obtaining a scoring result of the evaluation object based on the first data score value and the second data score value.

[0006] As an example, the step of scoring the network evaluation index data according to a preset scoring model and the benchmark value to obtain a first data score value includes: scoring the network evaluation index data based on the scoring model and the benchmark value to obtain a linear score value; testing the distribution of the linear score value to obtain a test result; if the test result does not meet a preset stopping condition, adjusting the preset parameters or updating the scoring model according to the test result, and returning to the step of scoring the network evaluation index data based on the scoring model and the benchmark value; if the test result meets the preset stopping condition, using the linear score value as the first data score value.

[0007] As one embodiment, the preset parameters include the probability values ​​of the lower limit range, the probability value of the full score range, the lower limit range, and the full score range, defined based on the minimum or maximum score benchmark value. Determining the benchmark value of the network evaluation index data based on the preset parameters includes: determining the probability values ​​of the lower limit range and the full score range corresponding to the network evaluation index data based on the cumulative distribution function value of the network evaluation index data; if the network evaluation index data is positive, determining the lower limit range of the network evaluation index data based on the probability value of the lower limit range. Based on the probability values ​​of the full score range of the network evaluation indicator data, the full score range of the network evaluation indicator data is determined, and the lower limit range and the full score range are used as the benchmark value of the network evaluation indicator data. If the network evaluation indicator data is negative, based on the probability values ​​of the lower limit range of the network evaluation indicator data, the full score range of the network evaluation indicator data is determined, and based on the probability values ​​of the full score range of the network evaluation indicator data, the lower limit range of the network evaluation indicator data is determined, and the lower limit range and the full score range are used as the benchmark value of the network evaluation indicator data.

[0008] As an example, the scoring model includes a first linear scoring model and a second linear scoring model. The step of scoring the network evaluation index data based on the scoring model and the benchmark value to obtain a linear score value includes: if the network evaluation index data is positive, substituting the benchmark value and the network evaluation index data into the first linear scoring model to obtain a first linear score value. The scoring criteria of the first linear scoring model include: if the network evaluation index data is less than or equal to the lower limit range, the score value is the lower limit; if the network evaluation index data is greater than or equal to the full score range, the score value is the full score; otherwise, the score value is the same as the lower limit range. The full score range is positively correlated with the lower limit score range and negatively correlated with the full score range. If the network evaluation index data is negative, the benchmark value and the network evaluation index data are substituted into the second linear scoring model to obtain a second linear score value. The scoring criteria of the second linear scoring model include: if the network evaluation index data is less than or equal to the lower limit score range, the score value is the full score value; if the network evaluation index data is greater than or equal to the full score range, the score value is the lower limit score value; otherwise, the score value is negatively correlated with the full score range and positively correlated with the lower limit score range. The linear score value is determined based on the first linear score value and the second linear score value.

[0009] As an example, it further includes: determining the scoring criteria of the first linear scoring model and / or the second linear scoring model when the network evaluation index data is greater than the lower limit of the scoring range but less than the full score range, based on the distance between the network evaluation index data and the predetermined minimum and maximum value ranges, respectively.

[0010] As an example, the distances between the network evaluation index data and the predetermined minimum and maximum value intervals are absolute distances, relative distances, or Minkowski distances.

[0011] As an example, the step of testing the distribution of the linear score values ​​to obtain the test result includes: testing the distribution of the linear score values ​​based on the SW test method or the KS test method to obtain the test result.

[0012] As an example, the nonlinear scoring model includes a first nonlinear scoring model and a second nonlinear scoring model. The step of scoring the network evaluation index data based on the preset nonlinear scoring model and the benchmark value to obtain a second data score value includes: if the network evaluation index data is positive, substituting the benchmark value and the network evaluation index data into the first nonlinear scoring model to obtain a first nonlinear score value. The scoring criteria of the first nonlinear scoring model include: if the network evaluation index data is less than or equal to the lower limit range, the score value is the lower limit; if the network evaluation index data is greater than or equal to the full score range, the score value is the full score; otherwise, the network evaluation index... The closer the data is to the lower limit range of the score, the faster the score increases. If the network evaluation index data is negative, the benchmark value and the network evaluation index data are substituted into the second nonlinear scoring model to obtain a second nonlinear score. The scoring criteria of the second nonlinear scoring model include: if the network evaluation index data is less than or equal to the lower limit range of the score, the score is the full score; if the network evaluation index data is greater than or equal to the full score range of the score, the score is the lower limit value; otherwise, the closer the network evaluation index data is to the full score range of the score, the faster the score increases. The second data score is determined based on the first nonlinear score and the second nonlinear score.

[0013] As an example, the method further includes: adjusting the first nonlinear scoring model and / or the second nonlinear scoring model according to preset first and second adjustment parameters when the network evaluation index data is greater than the lower limit of the score range but less than the full score range.

[0014] As an example, the nonlinear scoring model is an exponential distribution scoring model or a beta distribution scoring model.

[0015] Secondly, this application also provides a network evaluation index scoring device, including a data acquisition module, a benchmark value determination module, a first scoring module, a second scoring module, and a third scoring module.

[0016] The data acquisition module is used to acquire at least one network evaluation index data for the evaluation object.

[0017] The benchmark value determination module is used to determine the benchmark value of the network evaluation index data based on preset parameters, wherein the preset parameters are determined based on the probability density of the network evaluation index data.

[0018] The first scoring module is used to score the network evaluation index data according to a preset scoring model and the benchmark value if the network evaluation index data is linear data, and obtain a first data score value.

[0019] The second scoring module is used to score the network evaluation index data based on a preset nonlinear scoring model and the benchmark value if the network evaluation index data is nonlinear data, thereby obtaining a second data score value.

[0020] The third scoring module is used to obtain the scoring result of the evaluation object based on the first data scoring value and the second data scoring value.

[0021] Thirdly, this application also provides a method for determining network input energy efficiency profiles, including: obtaining the scoring results of the evaluation object; and determining the network input energy efficiency profile of the evaluation object based on the scoring results and the weights corresponding to each network evaluation index of the evaluation object.

[0022] Fourthly, this application also provides a network input energy efficiency profile determination device, including a scoring result acquisition module and a profile determination module.

[0023] The rating result acquisition module is used to obtain the rating results of the evaluation object.

[0024] The profile determination module is used to determine the network input efficiency profile of the evaluation object based on the scoring results and the weights corresponding to each network evaluation index of the evaluation object.

[0025] Fifthly, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement any of the network evaluation index scoring methods or the network input energy efficiency profile determination methods described above.

[0026] Sixthly, this application also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the network evaluation index scoring method or the network input energy efficiency profile determination method as described above.

[0027] Seventhly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the network evaluation index scoring method or the network input energy efficiency profile determination method as described above.

[0028] The method and apparatus for determining network evaluation index scoring and network input energy efficiency profile provided in this application determine the benchmark value of network evaluation index data by preset parameters, reducing the strong dependence of the scoring model on the maximum or minimum value. By adjusting the preset parameters and the scoring models of different data types, the scoring results can be adjusted, thus meeting the evaluation and processing needs of multiple types of network evaluation indicators. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a flowchart illustrating the scoring method for the network evaluation metrics provided in this application.

[0031] Figure 2a This is a frequency plot of data samples for a network evaluation index provided in this application. Figure 2b This is a frequency plot of data samples for another network evaluation metric provided in this application.

[0032] Figure 3a This is one of the schematic diagrams for determining the reference value provided in this application. Figure 3b This is the second schematic diagram for determining the benchmark value provided in this application.

[0033] Figure 4a This is one of the test results provided in this application. Figure 4b This is the second test result provided in this application. Figure 4c This is the third test result provided in this application. Figure 4d This is the fourth test result provided in this application.

[0034] Figure 5a This is one of the diagrams showing the comparison of test results provided in this application. Figure 5b This is the second diagram showing the comparison of test results provided in this application.

[0035] Figure 6 The scoring model provided in this application varies with parameters. r A diagram showing the comparison of score values ​​during the iterative optimization process.

[0036] Figure 7 This is one of the schematic diagrams illustrating the differences in score values ​​of the nonlinear scoring model provided in this application.

[0037] Figure 8 This is the second schematic diagram illustrating the differences in score values ​​of the nonlinear scoring model provided in this application.

[0038] Figure 9 This is the third illustration of the score value difference of the nonlinear scoring model provided in this application.

[0039] Figure 10 This is a schematic diagram of the network evaluation index scoring device provided in this application.

[0040] Figure 11 This is a flowchart illustrating the method for determining the network input energy efficiency profile provided in this application.

[0041] Figure 12a This is a schematic diagram of the network profile breakdown provided in this application; Figure 12b This is a schematic diagram illustrating the network input energy efficiency profile provided in this application; Figure 12c This is a schematic diagram of the comprehensive network profile provided in this application.

[0042] Figure 13 This is a schematic diagram of the network input energy efficiency profiling device provided in this application.

[0043] Figure 14 This is a schematic diagram of the network input energy efficiency profiling system provided in this application.

[0044] Figure 15 This is a schematic diagram of the structure of the electronic device provided in this application. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0046] It should be noted that all actions involving the acquisition of signals, information, or data in this application are carried out in compliance with the relevant data protection laws and policies of the country where the application is located, and with the authorization granted by the owner of the relevant device.

[0047] Figure 1 This is a flowchart illustrating the scoring method for network evaluation metrics provided in this application, as shown below. Figure 1 As shown, this application provides a network evaluation index scoring method, including the following steps S110-S150.

[0048] Step S110: Obtain network evaluation indicator data for the evaluation object; the evaluation object can be a network evaluation object with multiple dimensions such as customers, business volume, revenue and benefits, and the network evaluation indicator data is the data of the network evaluation indicators set in advance for each evaluation object at a preset time point or time period.

[0049] To improve the accuracy of network evaluation index scores, the network evaluation index data needs to be preprocessed after acquisition. Preprocessing includes handling outliers and default values.

[0050] Network evaluation metrics data can be represented as ,in, Indicates the first i A sample of network evaluation metrics data. n This represents the total number of samples.

[0051] Step S120: Determine the benchmark value of the network evaluation index data based on preset parameters, wherein the preset parameters are determined based on the probability density of the network evaluation index data; the preset parameters are parameters determined based on probability density on the basis of the existing minimum or maximum value as the scoring benchmark value.

[0052] Step S130: If the network evaluation index data is linear data, the network evaluation index data is scored according to a preset scoring model and the benchmark value to obtain a first data score value. Specifically, the scoring model can score the network evaluation index data by comparing it with the benchmark value, and based on the comparison result and the scoring criteria corresponding to the comparison result, to obtain the first data score value. The score obtained by the scoring model and the benchmark value on the network evaluation index data can be directly used as the first data score value, or the score can be verified, and the score that passes the verification is used as the first data score value.

[0053] Step S140: If the network evaluation index data is nonlinear data, the network evaluation index data is scored based on a preset nonlinear scoring model and the benchmark value to obtain a second data score value. Specifically, the nonlinear scoring model can score the network evaluation index data by comparing it with the benchmark value and according to the comparison result and the scoring standard corresponding to the comparison result to obtain a second data score value.

[0054] Step S150: Obtain the scoring result of the evaluation object based on the first data score and the second data score. If all network evaluation index data are linear data, the first data score is used as the scoring result; if all network evaluation index data are non-linear data, the second data score is used as the scoring result; if the network evaluation index data includes both linear and non-linear data, the combined first and second data score is used as the scoring result.

[0055] It is understandable that this application determines the baseline value of network evaluation index data by setting preset parameters, reducing the strong dependence of the scoring model on the maximum or minimum value. By adjusting the preset parameters and the scoring models of different data types, the scoring results can be adjusted, thus meeting the evaluation and processing needs of multiple types of network evaluation indicators.

[0056] Based on the above embodiments, as an optional embodiment, the preset parameters include the probability values ​​of the lower limit range of the scoring, the probability values ​​of the full score range, the lower limit range of the scoring, and the full score range of the scoring, defined based on the minimum or maximum value of the scoring benchmark value. a’ Defined as cumulative distribution function Data values ​​reached or exceeded alpha The corresponding threshold data value, at this time a’= It can be seen that the expression for the probability value of the lower limit range of the score is as follows: ,in, alpha For the corresponding lower limit range of scores a’ The set probability value. Similarly, the parameters... b’ Defined as representing the cumulative distribution function Data values ​​reached or exceeded 1-beta The corresponding threshold data value, at this time b’= It can be seen that the expression for the probability value of the range of full scores is: , of which 1- beta For the corresponding full score range b’ The set probability value.

[0057] Correspondingly, determining the benchmark value of the network evaluation index data based on preset parameters includes steps S121-S123.

[0058] Step S121: Based on the cumulative distribution function value of the network evaluation index data, determine the probability values ​​of the lower limit range and the full score range corresponding to the network evaluation index data. Specifically, draw a frequency histogram of the network evaluation index data, divide the data into several intervals (called bins or bars), and count the number / frequency of data points in each interval. Calculate the cumulative distribution function (CDF) of the network evaluation index data, as shown in the following formula:

[0059] ;

[0060] in, Indicates in total data n In, less than or equal to x Number of data points / frequency.

[0061] After determining the cumulative distribution function, set the initial probability values ​​for the network evaluation index data, that is, set the probability values ​​for the lower limit range of the scores. alpha Full score range 1-beta .

[0062] Step S122: If the network evaluation index data is positive, determine the lower limit range of the network evaluation index data based on the probability value of the lower limit range, and determine the full score range of the network evaluation index data based on the probability value of the full score range. Use the lower limit range and the full score range as the benchmark values ​​for the network evaluation index data. Specifically, "positive" network evaluation index data means that a larger value is better.

[0063] The process of determining the lower limit range of the network evaluation index data based on the probability values ​​of the lower limit range of the network evaluation index data is to find the range that satisfies the CDF function from the network evaluation index data. The data and the algorithm process expression are as follows: .

[0064] The process of determining the full score range of the network evaluation index data based on the probability values ​​of the full score range is to find the range that satisfies the CDF function from the network evaluation index data. The data and the algorithm process expression are as follows: .

[0065] Step S123: If the network evaluation index data is negative, determine the full score range of the network evaluation index data based on the probability value of the lower limit range of the network evaluation index data, and determine the lower limit range of the network evaluation index data based on the probability value of the full score range of the network evaluation index data. The lower limit range and the full score range are used as the benchmark values ​​for the network evaluation index data. Smaller values ​​for the network evaluation index data are generally considered better.

[0066] The process of determining the lower limit range of the network evaluation index data based on the probability values ​​of the full score range is equivalent to finding the range of values ​​that satisfy the CDF function from the network evaluation index data. The data and the algorithm process expression are as follows: .

[0067] The process of determining the full score range of the network evaluation index data based on the probability values ​​of the lower limit range of the network evaluation index data is to find the range of the full score range from the network evaluation index data using the CDF function. The data and the algorithm process expression are as follows: .

[0068] Figure 2a This is a frequency plot of data samples for a network evaluation index provided in this application. Figure 2b This is a frequency plot of data samples for another network evaluation metric provided in this application. Figure 2a This is a frequency map of data samples for 5G cell utilization data. The original 5G cell utilization data consisted of 1672 samples. To simplify the analysis, this embodiment of the application resampled the data according to the probability distribution to obtain 186 data samples. Figure 2b This is a frequency plot of broadband port utilization data. The original broadband port utilization data consisted of 222,210 samples, which were resampled to obtain 186 data samples. The data shows that the maximum values ​​of 5G cell utilization data are sparsely distributed, and the maximum value range is easily affected by individual values; the broadband port utilization data has a high density of near-zero values, which can easily influence the minimum value range.

[0069] Figure 3a This is one of the schematic diagrams for determining the reference value provided in this application. Figure 3b This is the second schematic diagram for determining the reference value provided in this application. This application uses... Figure 3a and Figure 3b Taking the broadband port utilization evaluation index as an example, this demonstrates the process of solving for the benchmark value. If we take parameters... alpha =0.002, parameter 1 -beta =0.95, then the lower limit range of the evaluation index score is... a’The score is 4.17, and the full score range for this evaluation indicator is 96.88. If the parameters are adjusted, the parameter... alpha =0.05, parameter 1-beta =0.93, then the lower limit range of the evaluation index score is... a’ The score was 20.83, and the full score range for this evaluation indicator was 92.19. Further optimization of the parameters requires further adjustments based on the test results of the scoring model.

[0070] Understandably, this application provides a scheme for determining parameterized benchmark values ​​based on probability density. From the CDF plot and frequency distribution plot, it can be seen that the determination of benchmark values ​​should avoid the influence of individual orphan data. The lower limit range and the full score range should be as smooth and continuous with the data distribution as possible in terms of probability density, so as to make the selection of benchmarks for network evaluation indicators more flexible, avoid the strong influence of individual values ​​on subsequent scoring models, and make the parameterized scoring model have a wider range of adjustment and more flexibility.

[0071] Based on the above embodiments, as an optional embodiment, the step of scoring the network evaluation index data according to the preset scoring model and the benchmark value to obtain the first data score value includes steps S131-S134.

[0072] Step S131: The network evaluation index data is scored based on the scoring model and the benchmark value to obtain a linear score value; specifically, the scoring model can be a piecewise linear scoring model, or the scoring model can be adjusted based on the benchmark value to make the adjustment range of the scoring model larger.

[0073] Step S132: The distribution of the linear score values ​​is tested to obtain the test result; specifically, it is determined whether the distribution of the linear score values ​​is a normal distribution, and the test result is whether it conforms to a normal distribution or does not conform to a normal distribution.

[0074] Step S133: If the test result does not meet the preset stopping condition, then adjust the preset parameters or update the scoring model according to the test result, and return to the step of scoring the network evaluation index data based on the scoring model and the benchmark value; specifically, the preset stopping condition is that the distribution of linear score values ​​conforms to a normal distribution or the maximum number of times the preset parameters are adjusted or the scoring model is updated is reached.

[0075] It should be noted that the scoring model in this application can be of various types. If the test results do not meet the preset stopping conditions, the preset parameters can be adjusted or a new scoring model can be updated for scoring.

[0076] Step S134: If the test result meets the preset stopping condition, the linear score value is used as the first data score value. If the test result conforms to a normal distribution or reaches the preset number of adjustments or updates, the linear score value is used as the first data score value.

[0077] It is understood that this application performs a test by judging the distribution of linear score values, and adjusts the preset parameters or updates the scoring model based on the test results until the test results meet the preset stopping conditions, which can effectively improve the accuracy of the first data score values.

[0078] Based on the above embodiments, as an optional embodiment, the scoring model includes a first linear scoring model and a second linear scoring model. The step of scoring the network evaluation index data based on the scoring model and the benchmark value to obtain a linear score value includes steps S1311-S1313. Based on the above embodiments, this application embodiment scores the network evaluation index data using a parameterized first linear scoring model and a second linear scoring model. By adjusting the preset parameters, the score values ​​of the network evaluation index data can be made to conform to a normal distribution as much as possible, thereby improving the comparability of network health assessments or profiles and ensuring good discriminative power in the assessment results.

[0079] Step S1311: If the network evaluation index data is positive, substitute the benchmark value and the network evaluation index data into the first linear scoring model to obtain a first linear score value. The scoring criteria of the first linear scoring model include: if the network evaluation index data is less than or equal to the lower limit range of the scoring, the score value is the lower limit value of the scoring; if the network evaluation index data is greater than or equal to the full score range of the scoring, the score value is the full score value of the scoring; otherwise, the score value is positively correlated with the full score range of the scoring and negatively correlated with the lower limit range of the scoring.

[0080] The expression for the first linear scoring model is shown below:

[0081] ;

[0082] in, a This is the lower limit of the score. b This is the maximum score. If the network evaluation index data is less than or equal to the lower limit of the score range... The rating value is the lower limit of the rating. a If the network evaluation index data is greater than or equal to the range of the maximum score, the score is the maximum score. b If the network evaluation index data is within the range The closer the score is a’ The smaller the score, the closer it is tob’ The higher the score.

[0083] Step S1312: If the network evaluation index data is negative, substitute the benchmark value and the network evaluation index data into the second linear scoring model to obtain a second linear score value. The scoring criteria of the second linear scoring model include: if the network evaluation index data is less than or equal to the lower limit range of the scoring, the score value is the full score value; if the network evaluation index data is greater than or equal to the full score range of the scoring, the score value is the lower limit value of the scoring; otherwise, the score value is negatively correlated with the full score range of the scoring and positively correlated with the lower limit range of the scoring.

[0084] The expression for the second linear scoring model is as follows:

[0085] ;

[0086] in, a This is the lower limit of the score. b The score is the maximum possible score. If the network evaluation index data is less than or equal to the lower limit of the score range, the score is the maximum possible score. b If the network evaluation index data is greater than or equal to the range of the full score, the score value is the lower limit of the score. a If the network evaluation index data is within the range The closer the score is a’ The higher the score, the closer to b’ The smaller the score.

[0087] Step S1313: Determine the linear score value based on the first linear score value and the second linear score value. If all network evaluation index data are positive, the first linear score value is used as the linear score value; if all network evaluation index data are negative, the second linear score value is used as the linear score value; if the network evaluation index data includes both positive and negative index data, the combined first and second linear score values ​​are used as the linear score value.

[0088] It is understandable that this application can adjust the linear score value through preset parameters to avoid the influence of the maximum or minimum value of a single indicator on the score result.

[0089] Based on the above embodiments, as an optional embodiment, the step of testing the distribution of the linear score values ​​to obtain the test results includes: testing the distribution of the linear score values ​​based on the SW test method or the KS test method to obtain the test results.

[0090] Let the linear score values ​​of the network evaluation index data be represented by the data sequence sample. xThis indicates that it is necessary to verify whether the data conforms to a normal distribution. For application scenarios where network evaluation index data has a discrete small sample size (sample size ≤ 5000), the Shapiro-Wilk test is used to verify whether it conforms to a normal distribution. The Wilk statistic can be calculated as shown in the following formula. W To obtain the data sequence samples corresponding to the linear score values ​​of the network evaluation index data. p value.

[0091] .

[0092] in, p The value is the minimum significance level that can reject the null hypothesis given a given sample of observations; The sample mean. n The number of samples in the data sequence. This is a constant related to the degree of fit of the normal distribution, which can be calculated using the corresponding formula; For the first i A data sequence sample.

[0093] The judgment criteria are: based on the given significance level and sample size. n To find the critical value, this embodiment selects... p A value ≥ 0.05 is the judgment condition, indicating that the data sequence sample conforms to a normal distribution. In practice, histograms, normal QQ plots and other methods can also be used to comprehensively evaluate the normality of linear score values.

[0094] In online profiling, if the sample size of the evaluation target's indicator data is very small (e.g., <50), the Kolmogorov-Smirnov (KS) test is a good non-parametric test method. It is performed by comparing the maximum difference between the cumulative distribution function of the sample and the cumulative distribution function of the normal distribution.

[0095] Table 1 illustrates the linear score values ​​and test results of the first and / or second linear scoring models during the iterative optimization process of adjusting preset parameters. Through iterative optimization of preset parameters, preset parameters that meet the discrimination requirements can be gradually obtained. The network evaluation index in Table 1 is named: 5G base station scale per 10,000 customers, unit: number of base stations / 10,000 customers. The sample size for the example is n=46. Assuming that this network evaluation index is a positive index, the larger the data value, the better. The score values ​​for different parameter combinations are shown in the table.

[0096] Table 1. Schematic diagram of the iterative optimization process of the scoring model.

[0097]

[0098] Figure 4aThis is one of the test results provided in this application. Figure 4b This is the second test result provided in this application. Figure 4c This is the third test result provided in this application. Figure 4d This is the fourth test result provided in this application. Figure 5a This is one of the diagrams showing the comparison of test results provided in this application. Figure 5b This is the second schematic diagram of the comparison of test results provided in this application. The iterative optimization process of the scoring model includes the following steps (11)-(13).

[0099] (11) Select initial alpha =0.01, 1-beta =0.95, as shown in rows 1-2 of Table 1, parameter a、b Given {60,100} and {40,100} respectively, the scale score for 5G base stations serving 10,000 customers is obtained. After the Swing test, the p-value is 0.047. Figure 4d As shown, if the SW test hypothesis H0≥0.05 is taken, the score value fails to meet the normality test.

[0100] (12) Parameter combination of the scoring model {(a lpha =0.01), 1-beta =0.95), a =30), b =100)}, as shown in row 3 of Table 1, this parameter combination { alpha, beta} Corresponding evaluation indicator lower limit range a' =28, the full score range for the evaluation indicators b' =89, which is the minimum and maximum value of this evaluation indicator. At this point, the score for the scale of 5G base stations per 10,000 customers is obtained. After the SW test, the p-value is <0.001. Figure 4b As shown, the score failed to meet the normality test.

[0101] (13) Continue the above process, the scoring model parameter combination {( alpha =0.01), 1-beta =0.90), a =60), b =100)}, as shown in row 9 of Table 1, at this time the scale score of 5G base stations for 10,000 customers is obtained. After SW test, the p-value is obtained as 0.056, as Figure 4a As shown, the score satisfies the normality test. Compared to the control group... Figure 5a and Figure 5b compared to, Figure 4a The linear score values ​​shown have good discriminative power. In this case, the linear score values ​​of the network evaluation index data of the evaluated object after verification can be used as the first data score value of the evaluated object.

[0102] It is worth noting that the parameter combinations obtained in the above iterative optimization process are not unique; multiple parameter combinations can be found, as shown in Table 1. Rows 9 and 10 both meet the SW test requirements and can be selected based on the actual situation of the evaluation index data. Secondly, the optimization iteration process can be implemented through optimization algorithms or optimization functions, which will not be described in detail in this application's embodiments. Thirdly, based on different data characteristics, the scoring model parameters are adjusted. alpha, 1-beta, a, b The sensitivity to changes varies, and parameters are combined based on the characteristics of the assessment data. In the above-mentioned parameter combinations for the scoring model, one set of { alpha 1-beta} The parameter corresponds to a set of { a'、b'} value, or possibly multiple sets of { alpha, 1-beta} parameters for the same { a'、b' The values, as shown in Table 1, do not affect the use of the scoring model in the embodiments of this application.

[0103] It is understood that this application provides a method for verifying scoring values. By verifying linear scoring values, the accuracy and discriminative power of linear scoring values ​​can be improved.

[0104] Based on the above embodiments, as an optional embodiment, the network evaluation index scoring method provided in this application further includes: determining the scoring criteria of the first linear scoring model and / or the second linear scoring model when the network evaluation index data is greater than the lower limit of the scoring range but less than the full score range, based on the distance between the network evaluation index data and the predetermined minimum value range and maximum value range respectively.

[0105] Based on the aforementioned first and second linear scoring models, this application proposes a method for evaluating network evaluation index data within a given range. The scoring optimization scheme is a distance-based parameterized scoring model. By defining multiple distance calculation methods with respect to the benchmark evaluation indicators, the evaluation model becomes more discriminative, and the network evaluation indicator scores can be adjusted more flexibly.

[0106] For the interval Based on the positive indicator data, the first linear scoring model can be further optimized to obtain the following model:

[0107] .

[0108] in, For network evaluation metrics data Distance from the minimum interval For network evaluation metrics data Distance from the range of maximum values.

[0109] For the interval For the negative index data, the second linear scoring model can be further optimized to obtain the following model:

[0110] .

[0111] in, For network evaluation metrics data Distance from the minimum interval For network evaluation metrics data Distance from the maximum value interval; network evaluation metric data Distance from the minimum interval The smaller the value, the higher the score, and vice versa.

[0112] To allow for more flexible adjustment of evaluation index scores, this invention provides several optional distance definitions, wherein the distances between the network evaluation index data and the predetermined minimum and maximum value intervals are absolute distances, relative distances, or Minkowski distances.

[0113] Absolute distance refers to network evaluation metrics data. The formulas for calculating the absolute distance distribution between the minimum and maximum value intervals are as follows: , .

[0114] Relative distance refers to network evaluation metrics data. The relative positional relationship between the data and the boundaries of the minimum and maximum value intervals is called relative distance. The relative distance refers to the relative distance between the data and the interval boundaries in the network evaluation. Within a certain range, for concentrated samples, the scores will be more sparse, which is conducive to the evaluation results having better discrimination. The formula for calculating the relative distance is as follows.

[0115] .

[0116] .

[0117] Minkowski distance (MD) is a general distance metric that can be adjusted based on specific circumstances. r value, r The parameter used to control the distance calculation method; for a one-dimensional index, the MD distance will not be affected by parameter 1 / r The MD distance varies depending on the changes, and the formula for calculating the MD distance is as follows.

[0118] .

[0119] .

[0120] To reflect the different degrees of difference among samples in network evaluation, this application uses the improved Minkowski distance solution described above. and It can better capture the relative position of samples in different intervals and different parameters. r The selected value can better reflect the discriminative characteristics of the scoring results.

[0121] Table 2 shows the scoring model based on distance optimization as a function of parameters. r Iterative optimization process diagram, assuming other parameters remain unchanged, only the parameters are optimized. r Different score values ​​for network evaluation index data are obtained, and the differences in these score values ​​are observed. The calculation process and implementation principle for improving the discrimination are compared. The network evaluation index in this embodiment is named: 4G equivalent single carrier frequency daily average carrying traffic, unit: GB / carrier frequency. The sample size in this embodiment is n=46. It is assumed that the network evaluation index is a positive index, that is, the larger value is better. The score value between the lower limit range of the evaluation index and the full score range of the index is obtained, where the distance is the Minkowski distance, which is less than... a' Interval, greater than b' For each interval, the first linear scoring model is used to obtain the corresponding linear score value.

[0122] Table 2. Scoring model varies with parameters r A diagram illustrating the iterative optimization process.

[0123]

[0124] Figure 6 The scoring model provided in this application varies with parameters. r A diagram illustrating the comparison of score values ​​during the iterative optimization process, showing how the scoring model changes with parameters. r The iterative optimization process includes the following steps (21)-(22).

[0125] (21) Suppose we have a set of parameters { alpha =0.01, 1-beta =0.90 ，a =40, b =100}, as shown in rows 1-5 of Table 1, through the parameter r Iterate to obtain different score values.

[0126] From Figure 6 Observations have determined that, under different parameters r Under these circumstances, different score values ​​are obtained for the average daily traffic carried by a 4G equivalent single carrier frequency; among them, parameters r =3, the obtained score is Figure 6 The bottom group of broken lines is the flattest, with the following parameters. r=2, the obtained score value is the second broken line from the bottom, and it can be clearly observed that the score value discrimination has improved, until r When the score is 0.2, the obtained score is the top line, where the score fluctuates the most; this change can also be shown by calculating the variance of the score for each group of parameters, with a variance of 152.95. r =3), variance 204.83 ( r =2), variance 337.54 ( r =1), variance 489.11 ( r =0.5), variance 685.88 ( r =0.2).

[0127] Furthermore, the optimization results can be selected using the aforementioned SW test method, as shown in Table 2, with parameters... r =2 (second row of Table 2), SW test p The value is 0.213, which satisfies the normality test; the scoring model can be further optimized by combining parameters. To determine better discriminative power in scoring, in practical applications, parameters with larger variances can also be selected, such as... Figure 6 Parameters in r =1 or r =0.5, even though the SW test does not satisfy a normal distribution at this point.

[0128] (22) Optimize by combining parameters, assuming another set of parameters { alpha =0.01, 1-beta =0.95 ，a =40, b =100}, as shown in rows 6-10 of Table 1, through the parameter r The process iterates to obtain different score values. The score values ​​obtained by optimizing this combination of parameters not only improve the "discrimination" but also significantly enhance the SW test. p Value, sixth row in Table 2 ( r =3), the seventh row in Table 2 ( r =2), the eighth row in Table 2 ( r =1) If all parameters satisfy the normality test, then the discriminative power and normality of the evaluation index scores can be balanced, and any combination of optimized parameters can be selected.

[0129] Understandably, this application makes the scoring model more discriminative by defining multiple distance calculation methods.

[0130] Based on the above embodiments, as an optional embodiment, the nonlinear scoring model includes a first nonlinear scoring model and a second nonlinear scoring model. The step of scoring the network evaluation index data based on the preset nonlinear scoring model and the benchmark value to obtain a second data score value includes steps S141-S143. For cases where the changes in network evaluation index data do not exhibit linear characteristics, in order to characterize the nonlinear characteristic that the closer the network evaluation index data is to the full score, the more difficult it is to improve, and the closer it is to the lower limit, the easier it is to improve, this application provides a nonlinear scoring model based on probability distribution.

[0131] Step S141: If the network evaluation index data is positive, substitute the benchmark value and the network evaluation index data into the first nonlinear scoring model to obtain a first nonlinear scoring value. The scoring criteria of the first nonlinear scoring model include: if the network evaluation index data is less than or equal to the lower limit range of the scoring, the scoring value is the lower limit value of the scoring; if the network evaluation index data is greater than or equal to the full score range of the scoring, the scoring value is the full score value of the scoring; otherwise, the closer the network evaluation index data is to the lower limit range of the scoring, the faster the scoring value increases.

[0132] If the network evaluation index data is within the range Between these points, the expression for the first nonlinear scoring model is shown below:

[0133] .

[0134] in, k These are parameters of the exponential distribution, used to adjust the shape of the exponential distribution; network evaluation metric data. The closer to the minimum value range a’ The faster the rating rises, the slower it rises.

[0135] Step S142: If the network evaluation index data is negative, substitute the benchmark value and the network evaluation index data into the second nonlinear scoring model to obtain a second nonlinear scoring value. The scoring criteria of the second nonlinear scoring model include: if the network evaluation index data is less than or equal to the lower limit range of the scoring, the scoring value is the full score value; if the network evaluation index data is greater than or equal to the full score range of the scoring, the scoring value is the lower limit value of the scoring; otherwise, the closer the network evaluation index data is to the full score range of the scoring, the faster the scoring value increases.

[0136] If the network evaluation index data is within the range Between these points, the expression for the second nonlinear scoring model is shown below:

[0137] .

[0138] in, k These are parameters of the exponential distribution, used to adjust the shape of the exponential distribution; network evaluation metric data. The closer to the full score range b’ The faster the rating rises, the slower it rises.

[0139] Step S143: Determine the second data score based on the first nonlinear score and the second nonlinear score. If all network evaluation index data are positive, the first nonlinear score is used as the second data score. If all network evaluation index data are negative, the second nonlinear score is used as the second data score. If the network evaluation index data includes both positive and negative index data, the combined first and second nonlinear score are used as the second data score.

[0140] Understandably, this application proposes a nonlinear scoring model that flexibly controls the shape of the index distribution through parameters to satisfy the nonlinear characteristics of various data and solve the problem that the changes in network evaluation indicators do not meet the linear law.

[0141] Based on the above embodiments, as an optional embodiment, the network evaluation index scoring method provided in this application further includes: adjusting the first nonlinear scoring model and / or the second nonlinear scoring model according to preset first adjustment parameters and second adjustment parameters when the network evaluation index data is greater than the lower limit of the scoring range but less than the full score range.

[0142] To adjust the shape and distribution of the nonlinear curve, two adjustment parameters are added. c, d The optimized expression for the first linear scoring model is shown below:

[0143] .

[0144] in, k It is a parameter of the exponential distribution, used to adjust the shape of the exponential distribution. c and d These are the first adjustment parameter and the second adjustment parameter, respectively, used to further adjust the shape of the mapping function.

[0145] Similarly, the optimized expression for the second linear scoring model is shown below:

[0146] .

[0147] in, k It is a parameter of the exponential distribution, used to adjust the shape of the exponential distribution. c and dThese are the first and second adjustment parameters, respectively, used to further adjust the shape of the mapping function. When c and d When all values ​​are 1, the nonlinear parameters only include k .

[0148] The aforementioned nonlinear scoring model provides a smooth transition, making the actual score value have a smaller impact on the score value when it is close to the upper limit, and a larger impact on the score value when it is close to the lower limit. This nonlinear mapping helps to balance the importance of different network evaluation indicators.

[0149] The following example uses a network evaluation index to illustrate the nonlinear scoring model. Specifically, the network evaluation index is the average daily traffic of a 5G cell, in GB, with a sample size of n=192. This index is assumed to be positive.

[0150] Figure 7 This is one of the schematic diagrams illustrating the score value differences of the nonlinear scoring model provided in this application, such as... Figure 7 As shown, the first nonlinear scoring model is used to score the network evaluation index data, assuming the parameters... k The data samples of network evaluation indicators range from 0.1 to 4. After sorting the data, the scores of the average daily traffic of 5G cells are stacked to obtain the results. Figure 7 The rating curve in the figure follows an exponential distribution, and the curve increases with the parameter. k The change reflects the curve of the network evaluation index data's score value from the lower limit to the full score, where... k The exponential curve exhibits its most significant characteristic when the value is 4.0, meaning it can characterize the evaluation index within the specified range. Nonlinear changes.

[0151] Figure 8 This is the second illustration of the score value difference in the nonlinear scoring model provided in this application, as shown below. Figure 8 As shown, the optimized first nonlinear scoring model is used to score the network evaluation index data, assuming... k and d Given a set of parameters, it is possible to measure c The impact of changes on the score of average daily traffic in 5G cells.

[0152] Figure 9 This is the third illustration of the score value difference in the nonlinear scoring model provided in this application, as shown below. Figure 9 As shown, the optimized first nonlinear scoring model is used to score the network evaluation index data, assuming... k and c Given a set of parameters, it is possible to measure d The impact of changes on the score of average daily traffic in 5G cells.

[0153] In practical applications, all of the above processes can be implemented through computer program instructions, with a wider variety of parameter combinations.

[0154] Optionally, the nonlinear scoring model is an exponential distribution scoring model or a beta distribution scoring model. The expression for the nonlinear scoring model based on the beta distribution scoring model is shown below:

[0155] .

[0156] in, It is the probability density function of the beta distribution, with parameters , It is the shape parameter of the beta distribution. This nonlinear scoring model is often used for scoring evaluation indicators of share-based and percentage-based networks.

[0157] It is understandable that this application uses parameters k , c , d This allows for precise control of the curve's shape, meeting the processing requirements of nonlinear indicators.

[0158] The network evaluation index scoring device provided in this application is described below. The network evaluation index scoring device described below can be referred to in correspondence with the network evaluation index scoring method described above.

[0159] Figure 10 This is a schematic diagram of the network evaluation index scoring device provided in this application, as shown below. Figure 10 As shown, this application also provides a network evaluation index scoring device, including a data acquisition module 1010, a benchmark value determination module 1020, a first scoring module 1030, a second scoring module 1040 and a third scoring module 1050.

[0160] The data acquisition module 1010 is used to acquire at least one network evaluation index data of the evaluation object.

[0161] The benchmark value determination module 1020 is used to determine the benchmark value of the network evaluation index data based on preset parameters, wherein the preset parameters are determined based on the probability density of the network evaluation index data.

[0162] The first scoring module 1030 is used to score the network evaluation index data according to a preset scoring model and the benchmark value if the network evaluation index data is linear data, and obtain a first data score value.

[0163] The second scoring module 1040 is used to score the network evaluation index data based on a preset nonlinear scoring model and the benchmark value if the network evaluation index data is nonlinear data, and obtain a second data score value.

[0164] The third scoring module 1050 is used to obtain the scoring result of the evaluation object based on the first data scoring value and the second data scoring value.

[0165] The network evaluation index scoring device provided in this application has the same technical effect as the network evaluation index scoring method, and will not be described in detail here.

[0166] The method for determining the network input energy efficiency profile provided in this application is described below. The method for determining the network input energy efficiency profile described below can be referred to in correspondence with the network evaluation index scoring method described above.

[0167] Figure 11 This is a flowchart illustrating the method for determining network input energy efficiency profiles provided in this application, such as... Figure 11 As shown, this application also provides a method for determining network input energy efficiency profiles, including the following steps S210-S220.

[0168] Step S210: Obtain the scoring results of the evaluation object. The evaluation object is divided into cities and prefectures. The scoring results of the evaluation object include the score values ​​corresponding to all network evaluation indicators of the evaluation object. The network evaluation indicators of the evaluation object are determined based on the network evaluation indicator system of the evaluation object, as shown in Table 3.

[0169] Table 3. Schematic diagram of the network evaluation index system for the evaluation objects.

[0170]

[0171]

[0172] Table 4. Suggested results of network evaluation indicators for the evaluated objects.

[0173]

[0174] It is worth mentioning that if a network evaluation index fails to obtain satisfactory normality test results or nonlinear scoring curves through the above optimization methods, the weight of that network evaluation index can be appropriately reduced so that the overall scoring results meet the expected evaluation requirements.

[0175] Step S220: Based on the scoring results and the weights corresponding to each network evaluation index of the evaluation object, determine the network input efficiency profile of the evaluation object.

[0176] Figure 12aThis is a schematic diagram of the network profile breakdown provided in this application; Figure 12b This is a schematic diagram illustrating the network input energy efficiency profile provided in this application; Figure 12c This is a schematic diagram of the comprehensive network profile provided in this application. The table below is the comprehensive score table for the network profile.

[0177] Table 5. Comprehensive Score Table for Online Profiles.

[0178]

[0179] Understandably, this application, based on the scoring method of network evaluation indicators, profiles the network input efficiency of different cities, which can improve the comparability of network evaluation indicators and make the evaluation results have good differentiation.

[0180] Figure 13 This is a schematic diagram of the network input energy efficiency profiling device provided in this application, such as... Figure 13 As shown, this application also provides a network input energy efficiency profile determination device, including a scoring result acquisition module 1310 and a profile determination module 1320.

[0181] The scoring result acquisition module 1310 is used to acquire the scoring results of the evaluation object.

[0182] The profile determination module 1320 is used to determine the network input efficiency profile of the evaluation object based on the scoring results and the weights corresponding to each network evaluation index of the evaluation object.

[0183] Figure 14 This is a schematic diagram of the network input energy efficiency profiling system provided in this application, such as... Figure 14 As shown, the network input energy efficiency profile determination system includes a data acquisition point 1410, a scoring result determination device 1420, and a profile device 1430.

[0184] Data acquisition point 1410 is used to acquire network evaluation index data based on the network evaluation index system of the evaluation object.

[0185] The scoring result determination device 1420 is used to select a scoring model and obtain network evaluation index data scoring results through preprocessing, scoring processing, data verification and parameter optimization.

[0186] The profiling device 1430 is used to obtain an input efficiency profile of the evaluated object based on the scoring results of the network evaluation index data and the weights corresponding to the network evaluation index.

[0187] Figure 15 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 15As shown, the electronic device may include a processor 1510, a communications interface 1520, a memory 1530, and a communication bus 1540. The processor 1510, communications interface 1520, and memory 1530 communicate with each other via the communication bus 1540. The processor 1510 can call logical instructions from the memory 1530 to execute the network evaluation index scoring method or the network energy efficiency profile determination method.

[0188] Furthermore, the logical instructions in the aforementioned memory 1530 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0189] On the other hand, this application also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the network evaluation index scoring method or the network input energy efficiency profile determination method provided by the above methods.

[0190] In another aspect, this application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the network evaluation index scoring method or the network input energy efficiency profile determination method provided by the above methods.

[0191] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0192] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0193] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A network evaluation metric scoring method, characterized in that, The method comprises: obtaining network evaluation index data of an evaluation object; determining a reference value of the network evaluation index data based on a preset parameter, wherein the preset parameter is determined based on a probability density of the network evaluation index data; if the network evaluation index data is linear data, scoring the network evaluation index data according to a preset scoring model and the reference value to obtain a first data score value; if the network evaluation index data is nonlinear data, scoring the network evaluation index data based on a preset nonlinear scoring model and the reference value to obtain a second data score value; obtaining a scoring result of the evaluation object according to the first data score value and the second data score value.

2. The network assessment metric scoring method of claim 1, wherein, The method comprises: scoring the network evaluation index data based on the scoring model and the reference value to obtain a linear score value; testing a distribution of the linear score value to obtain a test result; if the test result does not meet a preset stopping condition, adjusting the preset parameter or updating the scoring model according to the test result, and returning to the step of scoring the network evaluation index data based on the scoring model and the reference value; if the test result meets the preset stopping condition, taking the linear score value as the first data score value.

3. The network assessment indicator scoring method of claim 1 or 2, wherein, The preset parameter comprises a probability value of a scoring lower limit value interval, a probability value of a scoring full score value interval, the scoring lower limit value interval and the scoring full score value interval defined on the basis that a minimum value or a maximum value is a scoring reference value. The method comprises: determining the probability value of the scoring lower limit value interval and the probability value of the scoring full score value interval corresponding to the network evaluation index data according to a cumulative distribution function value of the network evaluation index data; if the network evaluation index data is positive index data, determining a scoring lower limit value interval of the network evaluation index data according to the probability value of the scoring lower limit value interval of the network evaluation index data, determining a scoring full score value interval of the network evaluation index data according to the probability value of the scoring full score value interval of the network evaluation index data, and taking the scoring lower limit value interval and the scoring full score value interval as the reference value of the network evaluation index data; 4. The network assessment metric scoring method of claim 3, wherein, if the network evaluation index data is negative index data, determining a scoring full score value interval of the network evaluation index data according to the probability value of the scoring lower limit value interval of the network evaluation index data, determining a scoring lower limit value interval of the network evaluation index data according to the probability value of the scoring full score value interval of the network evaluation index data, and taking the scoring lower limit value interval and the scoring full score value interval as the reference value of the network evaluation index data. The scoring model comprises a first linear scoring model and a second linear scoring model, and the method comprises: scoring the network evaluation index data based on the scoring model and the reference value to obtain a linear score value; If the network evaluation index data is positive index data, the baseline value and the network evaluation index data are substituted into the first linear scoring model to obtain a first linear scoring value, and the scoring standard of the first linear scoring model includes: if the network evaluation index data is less than or equal to the scoring lower limit value interval, the scoring value is the scoring lower limit value, if the network evaluation index data is greater than or equal to the scoring full score value interval, the scoring value is the scoring full score value, otherwise, the scoring value is positively correlated with the scoring full score value interval and negatively correlated with the scoring lower limit value interval; If the network evaluation index data is negative index data, the baseline value and the network evaluation index data are substituted into the second linear scoring model to obtain a second linear scoring value, and the scoring standard of the second linear scoring model includes: if the network evaluation index data is less than or equal to the scoring lower limit value interval, the scoring value is the scoring full score value, if the network evaluation index data is greater than or equal to the scoring full score value interval, the scoring value is the scoring lower limit value, otherwise, the scoring value is negatively correlated with the scoring full score value interval and positively correlated with the scoring lower limit value interval; The linear scoring value is determined according to the first linear scoring value and the second linear scoring value.

5. The network assessment metric scoring method of claim 4, wherein, Further comprising: According to the distance between the network evaluation index data and the pre-determined minimum value interval and maximum value interval, respectively, the scoring standard of the first linear scoring model and / or the second linear scoring model is determined when the network evaluation index data is greater than the scoring lower limit value interval and less than the scoring full score value interval.

6. The network assessment metric scoring method of claim 5, wherein, The distance between the network evaluation index data and the pre-determined minimum value interval and maximum value interval is absolute distance, relative distance or Minkowski distance.

7. The network assessment metric scoring method of claim 2, wherein, The distribution of the linear scoring value is tested to obtain a test result, including: The distribution of the linear scoring value is tested based on the SW test method or the KS test method to obtain the test result.

8. The network assessment metric scoring method of claim 3, wherein, The non-linear scoring model includes a first non-linear scoring model and a second non-linear scoring model, and the network evaluation index data is scored based on the pre-determined non-linear scoring model and the baseline value to obtain a second data scoring value, including: If the network evaluation index data is positive index data, the baseline value and the network evaluation index data are substituted into the first non-linear scoring model to obtain a first non-linear scoring value, and the scoring standard of the first non-linear scoring model includes: if the network evaluation index data is less than or equal to the scoring lower limit value interval, the scoring value is the scoring lower limit value, if the network evaluation index data is greater than or equal to the scoring full score value interval, the scoring value is the scoring full score value, otherwise, the closer the network evaluation index data is to the scoring lower limit value interval, the faster the scoring value rises. If the network evaluation index data is negative index data, the reference value and the network evaluation index data are substituted into the second nonlinear scoring model to obtain a second nonlinear scoring value, and the scoring standard of the second nonlinear scoring model comprises: if the network evaluation index data is less than or equal to the scoring lower limit value interval, the scoring value is a scoring full score value, if the network evaluation index data is greater than or equal to the scoring full score value interval, the scoring value is a scoring lower limit value, otherwise, the closer the network evaluation index data is to the scoring full score value interval, the faster the rising speed of the scoring value is; According to the first nonlinear scoring value and the second nonlinear scoring value, the second data scoring value is determined.

9. The network assessment metric scoring method of claim 8, wherein, Further comprising: According to the first adjustment parameter and the second adjustment parameter, when the network evaluation index data is greater than the scoring lower limit value interval and less than the scoring full score value interval, the first nonlinear scoring model and / or the second nonlinear scoring model is adjusted.

10. The network assessment metric scoring method of claim 8 or 9, wherein, The nonlinear scoring model is an exponential distribution scoring model or a beta distribution scoring model.

11. A network evaluation metric scoring apparatus, comprising: Comprising: A data acquisition module is configured to acquire at least one network evaluation index data of an evaluation object; A reference value determination module is configured to determine a reference value of the network evaluation index data based on a preset parameter, wherein the preset parameter is determined based on the probability density of the network evaluation index data; A first scoring module is configured to score the network evaluation index data according to a preset scoring model and the reference value to obtain a first data scoring value if the network evaluation index data is linear data; A second scoring module is configured to score the network evaluation index data based on a preset nonlinear scoring model and the reference value to obtain a second data scoring value if the network evaluation index data is nonlinear data; A third scoring module is configured to obtain a scoring result of the evaluation object according to the first data scoring value and the second data scoring value.

12. A network input energy efficiency portrait determination method based on the network evaluation index scoring method of any one of claims 1-10, characterized in that, Comprising: Obtaining a scoring result of an evaluation object; According to the scoring result and the weight corresponding to each network evaluation index of the evaluation object, a network input energy efficiency portrait of the evaluation object is determined.

13. A network energy efficiency profile determination apparatus, characterized by: A network input energy efficiency portrait determination method according to claim 12, comprising: A scoring result acquisition module is configured to acquire a scoring result of an evaluation object; A portrait determination module is configured to determine a network input energy efficiency portrait of the evaluation object according to the scoring result and the weight corresponding to each network evaluation index of the evaluation object.

14. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the network evaluation index scoring method according to any one of claims 1 to 10 or the network input energy efficiency portrait determination method according to claim 12.

15. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the network evaluation index scoring method according to any one of claims 1 to 10 or the network input energy efficiency portrait determination method according to claim 12.

16. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the network evaluation index scoring method according to any one of claims 1 to 10 or the network input energy efficiency portrait determination method according to claim 12.

Images