Regional development quality comprehensive evaluation method and device based on weight fusion and medium

Abstract

The invention discloses a regional development quality comprehensive evaluation method and device based on weight fusion and a medium, and relates to the field of regional development quality assessment, and the method comprises the steps: obtaining multi-dimensional evaluation index data of a to-be-evaluated region, carrying out the data statistics and standardization processing, and obtaining the processed multi-dimensional evaluation index data, an entropy weight method, a CRITIC method and an MAR-IDW method are respectively utilized to determine an entropy weight, a normalization weight and a linear normalization weight of each index; according to the entropy weight, the normalization weight and the linear normalization weight of each index, using an adaptive weight fusion strategy to determine a fusion weight of each index; and according to the fusion weight of each index, determining a comprehensive evaluation index of the to-be-evaluated region by using a TOPSIS method of improving positive and negative ideal solutions based on quantiles so as to determine the development quality of the to-be-evaluated region. According to the invention, more accurate comprehensive evaluation of regional development quality can be realized.

Description

Technical Field

[0001] This application relates to the field of regional development quality assessment, and in particular to a method, equipment, and medium for comprehensive evaluation of regional development quality based on weight fusion. Background Technology

[0002] With the advancement of high-quality regional development, scientifically assessing the quality of regional development has become a crucial task for governments and decision-making departments at all levels. The evaluation of regional development quality typically involves a multi-dimensional indicator system encompassing economic, social, and ecological aspects, with significant differences and strong correlations among these indicators. Traditional evaluation methods often rely on the max-min normalization method, which is susceptible to outliers, leading to large fluctuations and poor stability in the evaluation results. Meanwhile, existing objective weighting methods (such as entropy weighting and CRITIC methods) can partially reflect data distribution characteristics, but often focus only on a single dimension, such as indicator dispersion or correlation, making it difficult to comprehensively reveal the interrelationships and dynamic changes among indicators. Furthermore, when evaluation data exhibits multi-dimensional and non-linear characteristics, single-weight models are prone to bias, reducing the scientific rigor, reliability, and accuracy of the evaluation results. Therefore, there is an urgent need for a comprehensive weighting determination method that can integrate multiple objective weighting mechanisms, enhance computational stability, and adapt to complex regional data structures. This method aims to improve the accuracy and robustness of regional development quality evaluation, providing a more scientific basis for government decision-making and policy formulation. Summary of the Invention

[0003] The purpose of this application is to provide a method, equipment, and medium for comprehensive evaluation of regional development quality based on weighted fusion, so as to achieve a more accurate comprehensive evaluation of regional development quality.

[0004] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a comprehensive evaluation method for regional development quality based on weighted fusion, including: Obtain multi-dimensional evaluation indicator data for the region to be evaluated; the multi-dimensional evaluation indicator data includes indicators of economic development level, economic efficiency, economic structure, science and technology input intensity, education input intensity, human resource input intensity, science and technology output quantity, regional coordination, urban-rural industrial coordination, urban-rural dual coordination, urban-rural income coordination, industrial wastewater discharge rate, solid waste comprehensive utilization rate, air quality, foreign trade level, foreign investment utilization capacity, per capita disposable income, per capita consumption, number of doctors per 10,000 people, and public library collection per 100 people; the economic development level indicator includes per capita GDP; the economic efficiency indicator includes per capita fiscal revenue; the economic structure indicator includes the industrial structure upgrading index and the proportion of the tertiary industry in GDP; the science and technology input intensity indicator includes science and technology fiscal expenditure / public fiscal expenditure; the education input intensity indicator includes education expenditure / fiscal expenditure; the human resource input intensity indicator includes per capita GDP ... efficiency indicator includes per capita GDP; the economic efficiency indicator includes per capita GDP; the economic structure indicator includes the industrial structure upgrading index and the proportion of the tertiary industry in GDP; the science and technology input intensity indicator includes science and technology fiscal expenditure / public fiscal expenditure; the education input intensity indicator includes education expenditure / fiscal expenditure; the human resource input intensity indicator includes per capita GDP; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology The indicators include: number of university students; the quantity of scientific and technological output, including patent authorizations per 10,000 people; the regional coordination indicator, including per capita GDP of the city / national per capita GDP; the urban-rural industrial coordination indicator, including added value of the primary industry / GDP; the urban-rural dual coordination indicator, including the dual contrast index; the urban-rural income coordination indicator, including the ratio of per capita disposable income in rural and urban areas; the industrial wastewater discharge rate indicator, including industrial wastewater discharge / GDP; the solid waste comprehensive utilization rate indicator, including the comprehensive utilization rate of general industrial solid waste; the air quality indicator, including average PM2.5 concentration; the foreign trade level indicator, including total import and export volume / GDP; the foreign investment utilization capacity indicator, including actual utilized foreign investment / GDP; the per capita disposable income indicator, including per capita disposable income; the per capita consumption indicator, including total retail sales of consumer goods / total population; the number of physicians per 10,000 people, including the number of physicians / population per 10,000 people; and the public library collection per 100 people, including the public library collection / total population. The multi-dimensional evaluation index data is subjected to data statistics and standardization processing to obtain the processed multi-dimensional evaluation index data. Based on the processed multi-dimensional evaluation index data, the entropy weight, normalized weight, and linear normalized weight of each index are determined using the entropy weight method, CRITIC method, and MAR-IDW method, respectively. Based on the entropy weight, normalized weight, and linear normalized weight of each indicator, the fusion weight of each indicator is determined using an adaptive weight fusion strategy. Based on the fusion weights of each indicator, the TOPSIS method based on quantile-based improved positive and negative ideal solutions is used to determine the comprehensive evaluation index of the region to be evaluated, so as to determine the development quality of the region to be evaluated.

[0005] In one embodiment, the multi-dimensional evaluation index data is subjected to data statistics and standardization processing to obtain processed multi-dimensional evaluation index data, specifically including: The multi-dimensional evaluation index data are statistically analyzed to obtain the mean, standard deviation, coefficient of variation, 5%-95% quantile, 10%-90% quantile, and 2%-98% quantile for each index; the coefficient of variation is determined based on the standard deviation and the mean. Based on the coefficient of variation, 5%-95% quantile, 10%-90% quantile, and 2%-98% quantile of each indicator, the maximum normalization process is performed on each indicator to obtain the number of multidimensional indicators after processing.

[0006] In one embodiment, based on the processed multi-dimensional evaluation index data, the entropy weight value of each index is determined using the entropy weight method, specifically including: Calculate the normalized proportion of each indicator in the region to be evaluated; The information entropy of each indicator is determined based on the normalization ratio described for each indicator. Based on the information entropy of each indicator, determine the entropy weight of each indicator.

[0007] In one embodiment, based on the processed multi-dimensional evaluation index data, the normalized weights of each index are determined using the CRITIC method, specifically including: Based on the processed multi-dimensional evaluation index data, determine the correlation coefficient matrix; Based on the correlation coefficient matrix, the conflict between the indicators is determined; Based on the processed multi-dimensional evaluation index data and correlation coefficient matrix, the information content of each index is determined; The weights of each indicator are determined based on their conflict and information content. The weights of each indicator are normalized to obtain the normalized weights of each indicator.

[0008] In one embodiment, based on the processed multi-dimensional evaluation index data, the linear normalized weights of each index are determined using the MAR-IDW method, specifically including: Based on the processed multi-dimensional evaluation index data, an absolute ideal region for each index is constructed; the absolute ideal region is a set of evaluation region indices that reach the absolute ideal threshold for the corresponding index. Based on the absolute ideal region, determine the mean matrix of the conditional ideal region; Construct an ideal deviation matrix based on the mean matrix of the ideal region under the given conditions; The ideal deviation matrix is ​​summed by both rows and columns to obtain the row deviation and column deviation. The row deviation and the column deviation are combined to obtain the comprehensive deviation of each indicator; The overall deviation is linearly normalized to obtain the linearly normalized weights of each index.

[0009] In one embodiment, based on the entropy weights, normalized weights, and linear normalized weights of each indicator, an adaptive weight fusion strategy is used to determine the fusion weights of each indicator, specifically including: Based on the entropy weight, the normalized weight, and the linear normalized weight of each indicator, the product fusion weight of each indicator is determined. Based on the entropy weight, normalized weight and linear normalized weight of each indicator, the arithmetic mean weight of each indicator is determined. The coefficient of variation of the weights of each indicator is determined based on the entropy weights, normalized weights, and linear normalized weights of each indicator. Based on the weighted coefficient of variation of each indicator, determine the average coefficient of variation of all indicators; Based on the average coefficient of variation, a fusion coefficient is generated; The fusion weight of each index is determined based on the fusion coefficient, the product fusion weight, and the arithmetic mean weight.

[0010] In one embodiment, based on the fusion weights of each indicator, the TOPSIS method based on quantile-based improved positive and negative ideal solutions is used to determine the comprehensive evaluation index of the region to be evaluated, specifically including: Based on the processed multi-dimensional evaluation index data and the fusion weights of each index, a weighted decision matrix is ​​determined; Based on the weighted decision matrix, determine the positive and negative ideal solutions for each indicator; Determine the distance between the region to be evaluated and the positive ideal solution; Determine the distance between the region to be evaluated and the negative ideal solution; The comprehensive evaluation index of the region to be evaluated is determined based on the distance between the region to be evaluated and the positive ideal solution and the distance between the region to be evaluated and the negative ideal solution.

[0011] In one embodiment, it further includes: Based on the processed multi-dimensional evaluation index data and the fusion weight of each index, the weighted score of each index is determined to identify the development quality shortcomings of the region to be evaluated.

[0012] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the weighted fusion-based regional development quality comprehensive evaluation method described above.

[0013] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the weighted fusion-based comprehensive evaluation method for regional development quality described above.

[0014] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a method, device, and medium for comprehensive evaluation of regional development quality based on weighted fusion. The method acquires multi-dimensional evaluation index data of the region to be evaluated; performs data statistics and standardization on the multi-dimensional evaluation index data to obtain processed multi-dimensional evaluation index data; based on the processed multi-dimensional evaluation index data, the entropy weight, normalized weight, and linear normalized weight of each index are determined using the entropy weight method, the CRITIC method, and the MAR-IDW method, respectively; based on the entropy weight, normalized weight, and linear normalized weight of each index, an adaptive weighted fusion strategy is used to determine the fusion weight of each index; based on the fusion weight of each index, the TOPSIS method based on quantile-improved positive and negative ideal solutions is used to determine the comprehensive evaluation index of the region to be evaluated, thereby determining the development quality of the region. This application employs a quantile-improved normalization method, which can effectively suppress the influence of outliers and avoid evaluation errors caused by extreme values ​​or uneven data in traditional normalization methods. Next, this application employs a multi-dimensional weighting calculation method, integrating different weighting calculation models to effectively avoid the bias that may result from a single method. Furthermore, it adaptively adjusts the weights according to actual conditions, ensuring the rationality and objectivity of the weights. Moreover, by combining the TOPSIS method based on quantile-improved positive and negative ideal solutions, this application can simultaneously consider the actual performance of indicators during the evaluation process, thereby achieving a more accurate comprehensive evaluation of regional development quality. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A flowchart illustrating a comprehensive evaluation method for regional development quality based on weighted fusion, provided as an embodiment of this application; Figure 2 A flowchart illustrating the determination of index weights using the MAR-IDW method provided in an embodiment of this application; Figure 3 Logical relationship diagram for the three weight determination methods; Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0017] The technical solutions of the embodiments 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, and 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.

[0018] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] To address the problems existing in the background technology, this application proposes a novel regional development quality assessment technique based on a weighted fusion-based comprehensive evaluation method. First, a quantile-improved normalization method is employed. Compared to traditional methods, this method effectively suppresses the influence of outliers, avoiding assessment errors caused by extreme values ​​or uneven data in traditional normalization methods. Next, this application integrates different weight calculation models through a multivariate fusion weight calculation method, effectively avoiding the bias that may arise from a single method, and adaptively adjusting the weights according to actual conditions to ensure the rationality and objectivity of the weights. Furthermore, by combining the quantile-improved positive and negative ideal solution TOPSIS method, this application can simultaneously consider the actual performance of indicators during the evaluation process, thereby achieving a more accurate comprehensive evaluation of regional development quality.

[0020] Through this series of innovative methodological improvements, this application not only overcomes the shortcomings of traditional methods and improves the accuracy and comprehensiveness of the assessment, but also provides policymakers with a more scientific and reliable decision support tool.

[0021] In one exemplary embodiment, such as Figure 1 As shown, a comprehensive evaluation method for regional development quality based on weighted fusion is provided, including the following steps: S1: Obtain multi-dimensional evaluation indicator data for the region to be evaluated; the multi-dimensional evaluation indicator data includes indicators of economic development level, economic efficiency, economic structure, science and technology input intensity, education input intensity, human resource input intensity, science and technology output quantity, regional coordination, urban-rural industrial coordination, urban-rural dual coordination, urban-rural income coordination, industrial wastewater discharge rate, solid waste comprehensive utilization rate, air quality, foreign trade level, foreign investment utilization capacity, per capita disposable income, per capita consumption, number of doctors per 10,000 people, and public library collection per 100 people; the economic development level indicator includes per capita GDP; the economic efficiency indicator includes per capita fiscal revenue; the economic structure indicator includes the industrial structure upgrading index and the proportion of the tertiary industry in GDP; the science and technology input intensity indicator includes science and technology fiscal expenditure / public fiscal expenditure; the education input intensity indicator includes education expenditure / fiscal expenditure; the human resource input intensity indicator includes... The indicators include: the number of college students per 10,000 people; the quantity of scientific and technological output indicators include the number of patents granted per 10,000 people; the regional coordination indicators include the city's per capita GDP / national per capita GDP; the urban-rural industrial coordination indicators include the added value of the primary industry / GDP; the urban-rural dual coordination indicators include the dual contrast index; the urban-rural income coordination indicators include the ratio of per capita disposable income in rural and urban areas; the industrial wastewater discharge rate indicators include industrial wastewater discharge / GDP; the solid waste comprehensive utilization rate indicators include the comprehensive utilization rate of general industrial solid waste; the air quality indicators include the average concentration of PM2.5; the foreign trade level indicators include total import and export volume / GDP; the foreign investment utilization capacity indicators include actual utilized foreign investment / GDP; the per capita disposable income indicators include per capita disposable income; the per capita consumption indicators include total retail sales of consumer goods / total population; the number of physicians per 10,000 people indicators include the number of physicians / population per 10,000 people; and the public library collection per 100 people indicators include the public library collection / total population. In this embodiment, " / " represents division.

[0022] In this embodiment, based on literature review, expert surveys, and the availability of regional data, a regional development quality evaluation index system Y is first constructed, and then data is analyzed. Data collection. For example, primary indicators (dimensional level) include indicators of economic development level, economic efficiency, economic structure, intensity of science and technology investment, intensity of education investment, intensity of human resource investment, quantity of science and technology output, regional coordination, urban-rural industrial coordination, urban-rural dual coordination, urban-rural income coordination, industrial wastewater discharge rate, comprehensive utilization rate of solid waste, air quality, degree of foreign trade, foreign investment utilization capacity, per capita disposable income, per capita consumption, number of doctors per 10,000 people, and public library collection per 100 people, etc.; secondary indicators (factor level) include indicators such as per capita GDP and per capita fiscal revenue. The specific indicators include: industrial structure upgrading index and the proportion of the tertiary industry in GDP; science and technology fiscal expenditure / public fiscal expenditure; education expenditure / fiscal expenditure; number of college students per 10,000 people; number of patents granted per 10,000 people; per capita GDP of the city / per capita GDP of the country; added value of the primary industry / GDP; dual contrast index; ratio of per capita disposable income of rural and urban areas; industrial wastewater discharge / GDP; comprehensive utilization rate of general industrial solid waste; average PM2.5 concentration; total import and export volume / GDP; actual utilization of foreign capital / GDP; per capita disposable income; social retail sales of consumer goods / total population; number of doctors / number of people per 10,000 people; and public library collection / total population.

[0023] S2: Perform data statistics and standardization on the multi-dimensional evaluation index data to obtain the processed multi-dimensional evaluation index data.

[0024] In one embodiment, S2 specifically includes: S21: Perform data statistics on the multi-dimensional evaluation index data to obtain the mean, standard deviation, coefficient of variation, 5%-95% quantile, 10%-90% quantile, and 2%-98% quantile for each index; the coefficient of variation is determined based on the standard deviation and the mean.

[0025] In practical applications, the collected multi-dimensional evaluation index data are statistically analyzed, including the mean, standard deviation, coefficient of variation, 5%-95% quantiles, 10%-90% quantiles, and 2%-98% quantiles. The formula for calculating the coefficient of variation (CV) is: .

[0026] in, Let be the coefficient of variation of the j-th index.

[0027] S22: Based on the coefficient of variation, 5%-95% quantile, 10%-90% quantile and 2%-98% quantile of each indicator, perform maximum normalization on each indicator to obtain the number of multidimensional indicators after processing.

[0028] In practical applications, due to the dimensional differences in indicators such as economic and innovation dimensions related to regional development quality, standardization is a crucial step before comprehensive evaluation. Standardization not only helps transform indicators with different dimensions to a unified scale but also improves the comparability and contribution of each indicator in the evaluation model. Common indicator standardization methods include Z-score standardization, maximum-minimum normalization, and maximum value normalization. Because of its advantages in strong interpretability, computational simplicity, and high efficiency in multi-indicator comparisons, this embodiment is based on the maximum value normalization method. However, regional development quality indicators often exhibit significant uneven distribution and extreme values. Directly using the traditional maximum value normalization method is easily affected by the outlier amplification effect, thus weakening the comparability between indicators. Therefore, this application proposes an improved maximum value normalization method based on the coefficient of variation and quantiles to standardize indicator data such as regional development quality, which often exhibit uneven distribution. Its core principle is: through calculation... This method quantifies the relative dispersion of data and dynamically adjusts the quantile intervals used for normalization accordingly, handling extreme values ​​differently. The core logic is that higher dispersion indicates a greater likelihood of extreme outliers, requiring stricter intervals to enhance stability; conversely, lower dispersion indicates a more uniform data distribution, allowing for looser intervals to retain more effective variation information. This improvement effectively reduces the interference of outliers and enhances the robustness and discriminative power of the standardized data.

[0029] First, the relative dispersion of the indicator data is quantitatively assessed using the coefficient of variation, and the quantile interval used for normalization is dynamically adjusted accordingly.

[0030] according to The value determines the degree of dispersion of each indicator data, and the quantile interval is selected: high dispersion data ( ): Employing the 2%-98% quantile range effectively filters out extreme value interference and enhances data stability; Medium discrete data ( ): Employing a 10%-90% quantile range to balance stability and discriminative power; low discreteness data ( Using a 5%-95% quantile range, this approach appropriately preserves valid information and internal differences in the original data, enhancing data discriminability while minimizing the risk of extreme values. This mechanism improves data robustness by filtering quantile ranges, preventing outliers from distorting evaluation results, and optimizes discriminability for low-variability indicators, significantly enhancing the comparability between different indicators and making the normalization results more scientific and reliable.

[0031] Let Y = (y ij ) m×n This is the original data matrix. Wherein, For the i-th region (the region to be evaluated) indicator The data values. The normalization calculation formula is as follows: For positive indicators: Use formula Normalize.

[0032] if Greater than the upper quantile If , then it is directly normalized to 1.

[0033] For negative indicators (the smaller the value, the better), a two-step transformation method is used: The first step is forward normalization: .

[0034] if Greater than the upper quantile If , then it is directly normalized to 1.

[0035] Step 2: Linear transformation: .

[0036] The negative index normalization adopts a two-step linear transformation method, which reverses the evaluation direction while maintaining the linear relationship and value range of the original data. This effectively avoids information loss that may be caused by repeated normalization and enhances the consistency and comparability of index processing.

[0037] In the formula: For the i-th region (the region to be evaluated) indicator The standardized data values, after normalization, yield a standardized matrix X=(x ij ) m×n .

[0038] This method not only unifies the dimensions and directions of different types of indicators, ensuring their comparability in comprehensive evaluation, but also ensures that each indicator has better noise resistance and discriminability after standardization. At the same time, it fully preserves the distribution information and comparability of the data subjects, effectively improving the reliability and accuracy of data analysis.

[0039] S3: Based on the processed multi-dimensional evaluation index data, the entropy weight, normalized weight, and linear normalized weight of each index are determined using the entropy weight method, CRITIC method, and MAR-IDW method, respectively.

[0040] Common methods for calculating the weights of quantitative indicators include the entropy weight method, the coefficient of variation method, and the CRITIC method. However, a single calculation method often has certain limitations, which may lead to inaccurate or one-sided weight calculation results. To avoid the shortcomings of a single method, this application adopts three complementary weight calculation methods to jointly evaluate the weights of regional development quality indicators, thereby comprehensively leveraging the advantages of each method and ensuring more accurate and comprehensive calculation results.

[0041] In one embodiment, S31: Based on the processed multi-dimensional evaluation index data X ( For the processed multi-dimensional evaluation index data X The i-th region indicator to be evaluated Standardized data values, X The dimension is , This refers to the total number of regions (e.g., the number of cities). (This refers to the total number of indicators). Using the entropy weight method, the entropy weight value of each indicator is determined, specifically including: S311: Calculate the normalized proportion of each indicator in the region to be evaluated.

[0042] In practical applications, the information content of each indicator is calculated based on information entropy theory. First, the indicators... Calculate the first Normalization ratio for each region: .

[0043] in, For the first The normalized proportion of the j-th indicator in each region; m is the total number of regions.

[0044] S312: Determine the information entropy of each indicator based on the normalization ratio of each indicator.

[0045] In practical applications, the formula for calculating information entropy is as follows: .

[0046] in, Let be the information entropy of the j-th indicator.

[0047] S313: Determine the entropy weight of each indicator based on the information entropy of each indicator.

[0048] In practical applications, the formula for calculating entropy weights is as follows: .

[0049] in, Let be the entropy weight of the j-th indicator; the smaller the information entropy, the higher the entropy weight. Let be the information entropy of the k-th indicator, and n be the total number of indicators.

[0050] In one embodiment, S32: Based on the processed multi-dimensional evaluation index data X The normalized weights of each indicator are determined using the CRITIC method, specifically including: S321: Determine the correlation coefficient matrix based on the processed multi-dimensional evaluation index data.

[0051] In this embodiment, the correlation coefficient is used to measure the degree of information overlap between any two indicators. The stronger the correlation, the more information redundancy there is.

[0052] The formula for calculating the correlation coefficient is as follows: .

[0053] in, For the first The first indicator and the first The Pearson correlation coefficients of the indicators, with a range of values. ; , These are the standardized values ​​of the j-th and k-th indicators, respectively; , The first , The average of the standardized indicators.

[0054] This formula is used to calculate the pairwise relationships between all indicators. , build Order correlation coefficient matrix .

[0055] S322: Determine the conflict between indicators based on the correlation coefficient matrix.

[0056] In this embodiment, conflict reflects the degree of difference between the indicator and other indicators. The greater the difference, the more unique information the indicator provides, and the higher its weight should be. Since the positive or negative value of correlation only represents the direction of association and does not affect the degree of information overlap, it is necessary to take... The absolute value of.

[0057] The formula for calculating conflict is: .

[0058] in, For the first The first indicator and the first The absolute value of the Pearson correlation coefficient for each indicator is used to eliminate the directional influence and retain only the correlation strength. The smaller, The larger, The higher the value, the better the indicator. The stronger the conflict.

[0059] S323: Determine the information content of each indicator based on the processed multi-dimensional evaluation index data and correlation coefficient matrix.

[0060] In this embodiment, the amount of information combines the index's own dispersion (standard deviation) and information uniqueness (average overlap with other indices). The higher the dispersion and the lower the overlap, the more effective information the index contains.

[0061] (1) Calculate the standard deviation of the index : .

[0062] in, The larger the value, the more obvious the differences in the indicator's own values, and the richer the basic information.

[0063] (2) Calculate the average correlation coefficient of the indicators (Synchronous absolute value taking).

[0064] .

[0065] In accordance with conflict calculations, the absolute values ​​of correlation coefficients are taken and then averaged to reflect the indicators. The average information overlap strength with other indicators. The larger the size, the more redundant information there is, and the lower the proportion of effective information.

[0066] (3) Calculate the final information content .

[0067] .

[0068] Integrating discreteness and uniqueness, The larger the value, the more effective information the indicator contains.

[0069] S324: Determine the weight of each indicator based on its conflict and information content.

[0070] In this embodiment, the overall importance of the indicator is obtained by "conflictability × information content", and then normalized to obtain the final weight, ensuring that the sum of the weights is 1.

[0071] (1) Calculate the comprehensive evaluation value : .

[0072] (2) Normalization to find weights : .

[0073] in, Let J be the weight of the j-th index determined using the CRITIC method.

[0074] S325: Normalize the weights of each indicator to obtain the normalized weights of each indicator.

[0075] In this embodiment, the weights of all indicators are normalized so that the sum of all weights is 1: .

[0076] in, Let be the normalized weight of the j-th indicator.

[0077] Current mainstream objective weighting methods are mostly based on single-dimensional statistical logic. The entropy weighting method assigns weights solely based on the dispersion of indicator values, completely ignoring the correlation between indicators. While the CRITIC method considers conflicts and overlaps between indicators through correlation coefficients, its underlying assumption is a linear and symmetrical correlation. However, in multi-indicator systems with inherent coupling mechanisms, such as regional development, the influence between indicators is often asymmetric and conditional. For example, one indicator (such as "R&D intensity") may be a powerful driving factor—its excellent performance often accompanies a high overall system level; while another indicator (such as "technology transfer rate") may act as a persistent bottleneck—even if other indicators are excellent, it often performs poorly. This structural difference in the roles of "driving force" and "bottleneck" cannot be effectively measured by dispersion or symmetrical correlation.

[0078] Therefore, an objective weighting method called Multi-sample Absolute Region-based Ideal Deviation Weighting (MAR-IDW) is proposed to overcome the shortcomings of traditional weight determination methods in terms of structural interpretability and robustness. The overall idea of ​​the MAR-IDW method is as follows: First, the method introduces the concept of an "absolute ideal region," using a set of samples that reach a preset threshold to represent the ideal state of an indicator, rather than relying on a single ideal point, significantly enhancing robustness to data noise and outliers. Based on this, the conditional expectation matrix B is calculated to characterize the average performance of other indicators when a certain indicator reaches its ideal level, thus constructing the ideal deviation matrix P. Subsequently, P is summed row-wise and column-wise to obtain the row deviation and column deviation, reflecting the importance of the indicator as a "structural influencing factor" and a "system bottleneck factor," respectively. Finally, the two are combined by setting a fusion coefficient α, and the weights of each indicator are obtained after normalization.

[0079] Compared to traditional objective weighting methods based solely on dispersion or correlation, the MAR-IDW method not only possesses superior mathematical properties (such as translation invariance and regional robustness), but more importantly, its weighting results have clear managerial interpretability—the weight magnitude directly reflects the comprehensive contribution of the indicator in the evaluation system, encompassing both "influence" and "weaknesses." This method provides a new approach to determining weights for multi-indicator comprehensive evaluation that combines robustness, objectivity, and structural interpretability.

[0080] In one embodiment, S33: Based on the processed multi-dimensional evaluation index data X The MAR-IDW method is used to determine the linearly normalized weights of each indicator. The specific steps include: S331: Based on the processed multi-dimensional evaluation index data, construct the absolute ideal region for each index; the absolute ideal region is a set of evaluation region indexes that reach the absolute ideal threshold for the corresponding index.

[0081] S332: Determine the mean matrix of the conditional ideal region based on the absolute ideal region.

[0082] S333: Construct the ideal deviation matrix based on the mean matrix of the ideal region under the given conditions.

[0083] S334: Summate the ideal deviation matrix by row and column respectively to obtain the row deviation and column deviation.

[0084] S335: The row deviation and the column deviation are fused to obtain the comprehensive deviation of each indicator.

[0085] S336: Perform linear normalization on the comprehensive deviation to obtain the linear normalized weights of each index.

[0086] The calculation steps of the MAR-IDW method are as follows: Step 1: Construct the Absolute Ideal Region.

[0087] Set an absolute ideal threshold This is used to define the "high level" or "ideal" state of an indicator. For the j-th indicator, its absolute ideal region... Defined as the set of sample indices that reach this threshold on this metric: .

[0088] in, This is standardized data after normalization.

[0089] If no samples reach the threshold for a certain indicator (i.e.) If the value of the index is taken as the ideal region, then the process degenerates to taking the sample with the maximum value of the index. .

[0090] Note: when And the data has been normalized to In this case, the ideal region is composed of samples that achieve a perfect score on the indicator (the optimal value after normalization of each indicator data is 1); if there are no perfect score samples, then the region is composed of samples with the maximum value under the indicator.

[0091] Step 2: Calculate the Conditional Ideal-Region Mean Matrix.

[0092] Based on the absolute ideal region of each indicator, the conditional expected mean of all indicators is calculated, forming a conditional ideal region mean matrix. : .

[0093] Matrix elements The statistical significance of the indicator is: At a high level (i.e., the sample belongs to) Under the condition of ), the index The sample mean of the values ​​is denoted as . .

[0094] Step 3: Construct the Ideal Deviation Matrix.

[0095] In Under a scale representing a perfectly ideal (full score) state, the ideal deviation matrix is ​​defined. : .

[0096] It quantifies the degree to which "when indicator j is ideal, indicator k deviates from the ideal state on average (the optimal value after normalization of each indicator's data is 1)". Setting diagonal elements to zero avoids meaningless comparisons between the indicator and itself. The closer the value is to 0, the better the performance of indicator k is when indicator j is ideal.

[0097] Step 4: Calculate row and column deviations.

[0098] For the ideal deviation matrix Summing by row and column respectively yields two deviation vectors representing the importance of different structures.

[0099] Row deviation: The independent influence of the characterization index. The larger the value, the better. When the performance is excellent, other indicators still have significant shortcomings (far from the ideal value) and fail to effectively drive the coordinated improvement of other indicators. This indicates that the indicator has a high degree of independence and belongs to the strong constraint condition of the system. Therefore, it should be given a high weight.

[0100] Column deviation: The shortcomings of the characterization index. The larger the value, the better the performance of the indicator when other indicators perform well. The fact that this indicator is still lagging behind indicates that it is a stubborn bottleneck of the system and a shortcoming factor that restricts the overall development of the system. Therefore, it needs to be given special attention in the evaluation.

[0101] Step 5: Composite Deviation.

[0102] To balance the "impact" and "shortcomings" of the indicators, a fusion coefficient is introduced. Calculate the overall deviation of each indicator. : .

[0103] in, Adjust the emphasis on row and column deviations. Generally, one can take... To show that everyone is treated equally.

[0104] Step 6: Normalize to determine the final weights.

[0105] The comprehensive deviation vector After performing linear normalization, the MAR-IDW weights are obtained: .

[0106] in, is the linearly normalized weight of the j-th indicator.

[0107] Final weight It satisfies nonnegativity and normality.

[0108] The following concrete example will fully demonstrate the MAR-IDW method in τ abs The calculation process under the condition of =1 is presented to verify the feasibility and intuitiveness of the method.

[0109] 1. Data and parameter settings.

[0110] Consider an evaluation problem involving 5 samples and 4 indicators. After standardization processing using the method proposed in this application, the raw data yields a benefit-based decision matrix X (all indicator values ​​have been normalized to the interval (0, 1]): .

[0111] Parameter settings: Absolute ideal threshold: τ abs =1.0.

[0112] Fusion coefficient: =0.5 (equal weight fusion row bias and column bias).

[0113] 2. Calculation steps.

[0114] 2.1 Constructing an absolutely ideal region .

[0115] For each indicator j, find the sample set where its value is equal to 1 (i.e., it has reached the absolute ideal level): Indicator 1: Samples 3 and 4 take a value of 1, therefore .

[0116] Indicator 2: Sample 5 takes a value of 1, therefore .

[0117] Indicator 3: Sample 2 takes a value of 1, therefore .

[0118] Indicator 4: Sample 1 takes a value of 1, therefore .

[0119] 2.2 Calculate the mean matrix B of the ideal region under the given conditions.

[0120] Based on the ideal regions of each indicator, calculate the mean matrix B of the conditional ideal regions, where B... j This indicates that when index j reaches the ideal level (i.e., sample ∈ Under the condition of ), the average performance of index j.

[0121] Calculation process: j=1 ( ): Sample 3: (1.00, 0.59, 0.31, 0.78).

[0122] Sample 4: (1.00, 0.69, 0.58, 0.36).

[0123] Mean: B1 = (1.00, 0.64, 0.445, 0.57).

[0124] j=2( ): Sample 5: (0.23, 1.00, 0.69, 0.58).

[0125] Mean: B2 = (0.23, 1.00, 0.69, 0.58).

[0126] j=3 ( ): Sample 2: (0.34, 0.67, 1.00, 0.46).

[0127] Mean: B3 = (0.34, 0.67, 1.00, 0.46).

[0128] j=4 ( ): Sample 1: (0.81, 0.42, 0.27, 1.00).

[0129] Mean: B4 = (0.81, 0.42, 0.27, 1.00).

[0130] Conditional ideal region mean matrix B: .

[0131] 2.3 Construct the ideal deviation matrix P.

[0132] definition (When j≠k; diagonal P) jj =0), which represents "the average difference between index k and the ideal value (1) when index j reaches the ideal level".

[0133] Ideal deviation matrix P: .

[0134] 2.4 Calculate row and column deviations.

[0135] Line deviation L row (Reflecting the influence of the indicator): .

[0136] Therefore L row =(1.345, 1.50, 1.53, 1.50).

[0137] Column deviation L col (Reflecting the shortcomings of the indicators): .

[0138] Therefore L col =(1.62, 1.27, 1.595, 1.39).

[0139] 2.5 Calculate the overall deviation S j .

[0140] use =0.5 equal-weight fusion row bias and column bias: .

[0141] The calculation yields: .

[0142] Therefore, S = (1.4825, 1.385, 1.5625, 1.445).

[0143] 2.6 Normalization yields the final weights.

[0144] Total deviation: ΣS=5.875.

[0145] Normalized MAR-IDW weights: .

[0146] Therefore, the final weight vector is: .

[0147] S4: Based on the entropy weight, normalized weight and linear normalized weight of each indicator, determine the fusion weight of each indicator using an adaptive weight fusion strategy.

[0148] like Figure 3 As shown, the three types of weights complement each other across three dimensions: discreteness (entropy weight method), correlation (CRITIC method), and deviation (MAR-IDW method). In this embodiment, to fully utilize the advantages of the three weighting methods and achieve a dynamic balance among the three objective weighting methods, a coefficient of variation-driven adaptive weight fusion strategy is designed. This mechanism intelligently adjusts the proportions of the entropy weight method, CRITIC method, and MAR-IDW method in the final weight fusion by calculating the overall variation characteristics of the indicator system, thereby achieving adaptive adjustment in situations where indicators are relatively consistent or significantly different.

[0149] The product fusion weight is achieved by directly multiplying the weights of the three methods and normalizing them. It emphasizes "consensus first"—the fusion weight will only be significantly increased when all methods agree on the importance of the indicator. It has a strong penalty for extremely low weights of a certain method (e.g., when a certain weight is 0.1, the product result approaches 0). It is suitable for scenarios that require strict screening of core indicators.

[0150] The arithmetic average weighting method linearly averages the weights of the three methods, treating each method's judgment equally. The result lies between the minimum and maximum weights, is simple to calculate, and preserves reasonable discrepancies between methods.

[0151] The dynamic fusion mechanism intelligently adjusts the proportion of each component using the coefficient of variation (CV): first, it calculates the coefficient of variation of the weights of each individual indicator. Then calculate the average coefficient of variation for all indicators. This reflects the consistency of weights. Based on the magnitude of the average coefficient of variation, the hyperbolic tangent function is used. Generates the adaptive fusion coefficient 'a'. This function can... Mapping to the interval [0,1) achieves smooth, non-linear adjustment. When the indicator weights are highly consistent ( ), The arithmetic mean is the primary method to ensure simple and efficient fusion; when the differences in indicator weights are significant ( (larger) It focuses on product fusion to suppress extreme values ​​and achieve adaptive optimization of weight allocation, and is suitable for multi-source weight fusion scenarios such as complex system evaluation.

[0152] In one embodiment, S4 specifically includes: S41: Determine the product fusion weight of each indicator based on the entropy weight, the normalized weight, and the linear normalized weight of each indicator.

[0153] When it is necessary to ensure that all evaluation methods reach a consensus on the importance of the indicators, a product-based weighting method is used: .

[0154] in, The product fusion weight is the j-th indicator.

[0155] S42: Determine the arithmetic average weight of each indicator based on the entropy weight, the normalized weight, and the linear normalized weight of each indicator.

[0156] The arithmetic average weighting method linearly averages the weights of the three methods, treating each method's judgment equally. The result lies between the minimum and maximum weights, is simple to calculate, and preserves reasonable discrepancies between methods.

[0157] .

[0158] in, Let be the arithmetic average weight of the j-th indicator.

[0159] S43: Determine the coefficient of variation of the weights of each indicator based on the entropy weights, the normalized weights, and the linear normalized weights of each indicator.

[0160] Using formula Calculate the coefficient of variation of the weights for each indicator; where, Let be the coefficient of variation of the weight of the j-th indicator; Let be the standard deviation of the three weights for the j-th indicator; Let be the average of the three weights for the j-th indicator.

[0161] S44: Determine the average coefficient of variation for all indicators based on the weighted coefficients of variation for each indicator.

[0162] Using formula Calculate the average coefficient of variation for each indicator; where, is the average coefficient of variation.

[0163] S45: Generate the fusion coefficient based on the average coefficient of variation.

[0164] Using formula Generate fusion coefficients; where, This is the fusion coefficient.

[0165] S46: Determine the fusion weight of each index based on the fusion coefficient, the product fusion weight, and the arithmetic mean weight.

[0166] Based on the fusion coefficient, fusion product, and arithmetic average weight: .

[0167] in, The fusion weight of the j-th indicator.

[0168] The advantages of this fusion mechanism are: When the three weights are relatively consistent ( (Smaller), tends to use arithmetic average weighting.

[0169] When there is a clear conflict in weights ( (For larger values), it tends to use product-fusion weights to effectively suppress extreme weight values.

[0170] S5: Based on the fusion weights of each indicator, the TOPSIS method based on quantile-based improved positive and negative ideal solutions is used to determine the comprehensive evaluation index of the region to be evaluated, so as to determine the development quality of the region to be evaluated.

[0171] In this embodiment, based on the quantile-improved index normalization matrix, this application proposes a TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method based on quantile-improved positive and negative ideal solutions. The TOPSIS method is widely used in comprehensive evaluation and decision-making problems. The basic principle of the TOPSIS method is to select the solution closest to the ideal solution (positive ideal solution) and furthest from the negative ideal solution. However, traditional TOPSIS methods typically use the maximum and minimum values ​​as the ideal and negative ideal solutions, respectively. This approach is easily affected by extreme data when dealing with data containing outliers or skewed distributions, thus impacting the reliability of the decision. To address this issue, this application proposes a TOPSIS method based on quantile-improved positive and negative ideal solutions. In the improved method, the ideal and negative ideal solutions are calculated from the upper and lower quantiles of each index and their corresponding weights, respectively. By using quantiles instead of maximum and minimum values, the interference of extreme values ​​can be avoided more effectively, enhancing the model's adaptability and robustness in practical decision-making.

[0172] In one embodiment, based on the fusion weights of each indicator, the TOPSIS method based on quantile-based improved positive and negative ideal solutions is used to determine the comprehensive evaluation index of the region to be evaluated, specifically including: S51: Determine the weighted decision matrix based on the processed multi-dimensional evaluation index data and the fusion weight of each index.

[0173] In this embodiment, the index normalization matrix (processed multi-dimensional evaluation index data) is based on quantile-improved data and the fusion weights of each index. Calculate the weighted decision matrix: .

[0174] The weighted decision matrix is ; This is the weighted value of the j-th indicator in the i-th region.

[0175] S52: Based on the weighted decision matrix, determine the positive and negative ideal solutions for each index.

[0176] In practical applications, positive and negative ideal solutions are determined based on the index normalization matrix improved by quantiles.

[0177] Positive ideal solution (optimal solution) The ideal value for each metric is derived from the maximum value of the normalized matrix based on quantile improvements, calculated in conjunction with the fused weights. It's important to note that this maximum value differs from the result in the traditional TOPSIS method, which is based on the maximum value of the normalized matrix and the weighted average. That is: .

[0178] Negative ideal solution (worst solution) The ideal value for each metric is calculated by taking the minimum value corresponding to the normalized matrix based on quantile improvement and the fused weights. It's important to note that this minimum value is not the same as the result of weighting the minimum value of the normalized matrix and the weights as in the traditional TOPSIS method. That is: .

[0179] Compared to the traditional TOPSIS method, the positive and negative ideal solutions based on quantiles are less affected by extreme data or outliers.

[0180] S53: Determine the distance between the region to be evaluated and the positive ideal solution.

[0181] S54: Determine the distance between the region to be evaluated and the negative ideal solution.

[0182] In this embodiment, Euclidean distance is typically used to measure distance: No. Distance between each region and the ideal solution : .

[0183] No. Distance between each region and the negative ideal solution : .

[0184] S55: Determine the comprehensive evaluation index of the region to be evaluated based on the distance between the region to be evaluated and the positive ideal solution and the distance between the region to be evaluated and the negative ideal solution.

[0185] In this embodiment, the relative closeness (comprehensive evaluation index) is calculated.

[0186] Relative closeness Indicates the first The formula for calculating the degree of closeness of each region to the ideal solution is: .

[0187] nature: .

[0188] The bigger the better: The closer the value is to 1, the closer the region is to the ideal solution, and the better the overall performance. Corresponding ideal solution, The corresponding negative ideal solution. According to The size is used to analyze the merits and demerits of all evaluation objects.

[0189] In one embodiment, it further includes: S6: Based on the processed multi-dimensional evaluation index data and the fusion weight of each index, determine the weighted score of each index to identify the development quality shortcomings of the region to be evaluated.

[0190] In this embodiment, the diagnosis of shortcomings in regional development quality is also included.

[0191] Calculate the weighted score for each indicator and identify the indicator with the lowest score as the current weakness: .

[0192] In this embodiment, the regional development quality comprehensive evaluation method based on weight fusion is implemented using software modules, the main software modules of which include: Data preprocessing module: Data statistical analysis and data normalization functions.

[0193] Weight calculation module: The weights of indicators are calculated using three methods: entropy weight method, CRITIC and MAR-IDW.

[0194] Weight fusion module: Adaptive weight fusion.

[0195] TOPSIS Evaluation: A comprehensive evaluation of TOPSIS based on quantile-improved positive and negative solutions.

[0196] Visualization output module: Provides ranking charts of regional development quality scores; radar charts of scores for various dimensions of regional development, etc.

[0197] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-described weighted fusion-based comprehensive evaluation method for regional development quality.

[0198] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the aforementioned weighted fusion-based comprehensive evaluation method for regional development quality.

[0199] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the aforementioned weighted fusion-based comprehensive evaluation method for regional development quality.

[0200] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 4As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media to run. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a weighted fusion-based comprehensive evaluation method for regional development quality.

[0201] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0202] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0203] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0204] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0205] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0206] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A comprehensive evaluation method for regional development quality based on weighted fusion, characterized in that, include: Obtain multi-dimensional evaluation indicator data for the region to be evaluated; the multi-dimensional evaluation indicator data includes indicators of economic development level, economic efficiency, economic structure, science and technology input intensity, education input intensity, human resource input intensity, science and technology output quantity, regional coordination, urban-rural industrial coordination, urban-rural dual coordination, urban-rural income coordination, industrial wastewater discharge rate, solid waste comprehensive utilization rate, air quality, foreign trade level, foreign investment utilization capacity, per capita disposable income, per capita consumption, number of doctors per 10,000 people, and public library collection per 100 people; the economic development level indicator includes per capita GDP; the economic efficiency indicator includes per capita fiscal revenue; the economic structure indicator includes the industrial structure upgrading index and the proportion of the tertiary industry in GDP; the science and technology input intensity indicator includes science and technology fiscal expenditure / public fiscal expenditure; the education input intensity indicator includes education expenditure / fiscal expenditure; the human resource input intensity indicator includes per capita GDP ... efficiency indicator includes per capita GDP; the economic efficiency indicator includes per capita GDP; the economic structure indicator includes the industrial structure upgrading index and the proportion of the tertiary industry in GDP; the science and technology input intensity indicator includes science and technology fiscal expenditure / public fiscal expenditure; the education input intensity indicator includes education expenditure / fiscal expenditure; the human resource input intensity indicator includes per capita GDP; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology output; the economic efficiency indicator includes science and technology The indicators include: number of university students; the quantity of scientific and technological output, including patent authorizations per 10,000 people; the regional coordination indicator, including per capita GDP of the city / national per capita GDP; the urban-rural industrial coordination indicator, including added value of the primary industry / GDP; the urban-rural dual coordination indicator, including the dual contrast index; the urban-rural income coordination indicator, including the ratio of per capita disposable income in rural and urban areas; the industrial wastewater discharge rate indicator, including industrial wastewater discharge / GDP; the solid waste comprehensive utilization rate indicator, including the comprehensive utilization rate of general industrial solid waste; the air quality indicator, including average PM2.5 concentration; the foreign trade level indicator, including total import and export volume / GDP; the foreign investment utilization capacity indicator, including actual utilized foreign investment / GDP; the per capita disposable income indicator, including per capita disposable income; the per capita consumption indicator, including total retail sales of consumer goods / total population; the number of physicians per 10,000 people, including the number of physicians / population per 10,000 people; and the public library collection per 100 people, including the public library collection / total population. The multi-dimensional evaluation index data is subjected to data statistics and standardization processing to obtain the processed multi-dimensional evaluation index data. Based on the processed multi-dimensional evaluation index data, the entropy weight, normalized weight, and linear normalized weight of each index are determined using the entropy weight method, CRITIC method, and MAR-IDW method, respectively. Based on the entropy weight, normalized weight, and linear normalized weight of each indicator, the fusion weight of each indicator is determined using an adaptive weight fusion strategy. Based on the fusion weights of each indicator, the TOPSIS method based on quantile-based improved positive and negative ideal solutions is used to determine the comprehensive evaluation index of the region to be evaluated, so as to determine the development quality of the region to be evaluated.

2. The regional development quality comprehensive evaluation method based on weighted fusion according to claim 1, characterized in that, The multi-dimensional evaluation index data is subjected to data statistics and standardization processing to obtain processed multi-dimensional evaluation index data, specifically including: The multi-dimensional evaluation index data are statistically analyzed to obtain the mean, standard deviation, coefficient of variation, 5%-95% quantile, 10%-90% quantile, and 2%-98% quantile for each index; the coefficient of variation is determined based on the standard deviation and the mean. Based on the coefficient of variation, 5%-95% quantile, 10%-90% quantile, and 2%-98% quantile of each indicator, the maximum normalization process is performed on each indicator to obtain the number of multidimensional indicators after processing.

3. The method for comprehensive evaluation of regional development quality based on weighted fusion according to claim 1, characterized in that, Based on the processed multi-dimensional evaluation index data, the entropy weight value of each index is determined using the entropy weight method, specifically including: Calculate the normalized proportion of each indicator in the region to be evaluated; The information entropy of each indicator is determined based on the normalization ratio described for each indicator. Based on the information entropy of each indicator, determine the entropy weight of each indicator.

4. The regional development quality comprehensive evaluation method based on weighted fusion according to claim 1, characterized in that, Based on the processed multi-dimensional evaluation index data, the normalized weights of each index are determined using the CRITIC method, specifically including: Based on the processed multi-dimensional evaluation index data, determine the correlation coefficient matrix; Based on the correlation coefficient matrix, the conflict between the indicators is determined; Based on the processed multi-dimensional evaluation index data and correlation coefficient matrix, the information content of each index is determined; The weights of each indicator are determined based on their conflict and information content. The weights of each indicator are normalized to obtain the normalized weights of each indicator.

5. The regional development quality comprehensive evaluation method based on weighted fusion according to claim 1, characterized in that, Based on the processed multi-dimensional evaluation index data, the linear normalized weights of each index are determined using the MAR-IDW method, specifically including: Based on the processed multi-dimensional evaluation index data, an absolute ideal region for each index is constructed; the absolute ideal region is a set of evaluation region indices that reach the absolute ideal threshold for the corresponding index. Based on the absolute ideal region, determine the mean matrix of the conditional ideal region; Construct an ideal deviation matrix based on the mean matrix of the ideal region under the given conditions; The ideal deviation matrix is ​​summed by both rows and columns to obtain the row deviation and column deviation. The row deviation and the column deviation are combined to obtain the comprehensive deviation of each indicator; The overall deviation is linearly normalized to obtain the linearly normalized weights of each index.

6. The method for comprehensive evaluation of regional development quality based on weighted fusion according to claim 1, characterized in that, Based on the entropy weight, normalized weight, and linear normalized weight of each indicator, an adaptive weight fusion strategy is used to determine the fusion weight of each indicator, specifically including: Based on the entropy weight, the normalized weight, and the linear normalized weight of each indicator, the product fusion weight of each indicator is determined. Based on the entropy weight, normalized weight and linear normalized weight of each indicator, the arithmetic mean weight of each indicator is determined. The coefficient of variation of the weights of each indicator is determined based on the entropy weights, normalized weights, and linear normalized weights of each indicator. Based on the weighted coefficient of variation of each indicator, determine the average coefficient of variation of all indicators; Based on the average coefficient of variation, a fusion coefficient is generated; The fusion weight of each index is determined based on the fusion coefficient, the product fusion weight, and the arithmetic mean weight.

7. The method for comprehensive evaluation of regional development quality based on weighted fusion according to claim 1, characterized in that, Based on the fusion weights of each indicator, the TOPSIS method based on quantile-based improved positive and negative ideal solutions is used to determine the comprehensive evaluation index of the region to be evaluated, specifically including: Based on the processed multi-dimensional evaluation index data and the fusion weights of each index, a weighted decision matrix is ​​determined; Based on the weighted decision matrix, determine the positive and negative ideal solutions for each indicator; Determine the distance between the region to be evaluated and the positive ideal solution; Determine the distance between the region to be evaluated and the negative ideal solution; The comprehensive evaluation index of the region to be evaluated is determined based on the distance between the region to be evaluated and the positive ideal solution and the distance between the region to be evaluated and the negative ideal solution.

8. The method for comprehensive evaluation of regional development quality based on weighted fusion according to claim 1, characterized in that, Also includes: Based on the processed multi-dimensional evaluation index data and the fusion weight of each index, the weighted score of each index is determined to identify the development quality shortcomings of the region to be evaluated.

9. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the weighted fusion-based comprehensive evaluation method for regional development quality according to any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the weighted fusion-based comprehensive evaluation method for regional development quality as described in any one of claims 1-8.

Images