Intelligent industrial operation analysis method and system

Abstract

The invention discloses an intelligent industrial operation analysis method and system, and relates to the field of big data analysis. Comprising the steps of 1, directionally processing data based on business requirements to obtain index characteristic variables for analysis, 2, respectively establishing an analysis model of government affair data and an analysis model of public data according to the index characteristic variables, and 3, analyzing the industrial development condition of a local area according to national data labels through the analysis model of the government affair data. And outputting a strong, medium or weak development grade according to the industrial development condition, analyzing the industrial ranking of the local area in the nationwide area through an analysis model of public data, and judging the grade according to the quantile of the local area ranking.

Description

Technical Field

[0001] The present invention discloses an intelligent industrial operation analysis method and system, and relates to the field of big data analysis. Background Art

[0002] The digital economy is the main economic form after the agricultural economy and the industrial economy. It takes digital industrialization and industrial digitalization as its main line, builds a support system, improves the governance system, promotes digital technology innovation, cultivates the data factor market, promotes the deep integration of digital technology and the real economy, and empowers the development of a digital society.

[0003] Industrial operation is an extremely complex system. From the overall perspective of promoting high-quality economic development, it is an important foundation to coordinate and plan a comprehensive analysis engine. However, there are some problems at present, such as:

[0004] The data of various industries are complex and numerous, the data of different industries vary, and the data quality stability varies. There is no intelligent and standardized industry operation analysis method to monitor and evaluate industry operations and reflect problems in the industry development process in order to assist decision-making. Summary of the invention

[0005] In response to the problems of the prior art, the present invention provides an intelligent industrial operation analysis method and system, which processes and manages market public data and government data, conducts monitoring and analysis, high-quality development evaluation, and regional benchmarking analysis, introduces dimensional indicators such as operating performance, development potential, risk resistance, social contribution, and green development, and introduces an evaluation that characterizes the industrial chain based on public enterprise data, so as to make a judgment on the development status of the regional macro-industry, monitor, analyze, warn, and forecast, and promptly discover problems in the process of regional economic development, thereby providing support for regional industrial development and decision-making.

[0006] The specific scheme proposed by the present invention is:

[0007] The present invention also provides an intelligent industrial operation analysis method, comprising:

[0008] Step 1: Perform targeted data processing based on business requirements to obtain indicator characteristic variables for analysis: perform data cleaning, binning and standardization on public data and government data respectively, and confirm indicator characteristic variables that can be used for model establishment. The indicator characteristic variables confirmed based on government data include: operating performance indicators, development capability indicators, risk resistance indicators, social contribution indicators and green development indicators. The indicator characteristic variables confirmed based on public data include industry scale indicators, operating growth indicators, personnel scale indicators and innovation capability indicators.

[0009] Step 2: Establish analysis models for government data and public data based on indicator characteristic variables.

[0010] Step 3: Analyze the local regional industrial development situation through the analysis model of government data according to national data labels, and output the development level as strong, medium or weak according to the industrial development situation.

[0011] The industry ranking of the local region within the national region is analyzed through an analysis model of public data, and the level is determined according to the percentile of the local region ranking.

[0012] Furthermore, when obtaining the indicator characteristic variables for analysis in step 1 of the intelligent industrial operation analysis method, it includes: using the univariate indicator analysis method to identify the characteristic variables with predictive ability for the local regional industry, applying the Pearson correlation coefficient to test the degree of correlation between each related characteristic variable, reducing the collinearity between the characteristic variables through the principal component analysis method, reducing the statistical dimension, and obtaining the final indicator characteristic variable.

[0013] Furthermore, in step 2 of the intelligent industry operation analysis method, an analysis model of government affairs data is established, including: the logical formula for establishing the model is as follows:

[0014]

[0015] w i It is the weight coefficient of each indicator characteristic variable in the overall model f. The initial weight coefficient of the model is preset, and the published digital economy, regional economic indicators, and ranking data are used as the basis for weight system correction to adjust the weight.

[0016] Diff represents the judgment of model f on a specific dimension With country data label y j Be consistent.

[0017] Furthermore, in step 2 of the intelligent industrial operation analysis method, an analysis model for public data is established based on a logistic regression algorithm, a random forest algorithm, a Bagging algorithm, an AdaBoost algorithm, a voting model algorithm, and a neural network algorithm.

[0018] Furthermore, in step 3 of the intelligent industry operation analysis method, the analysis model of public data is used to define a national target library of regions of the same level according to the local regional level to be assessed as needed.

[0019] For each target in the regional target library, six models are used for analysis;

[0020] Sort the models by their AUC indicators, assign each model a weight of 25%, 20%, 20%, 15%, 10%, and 10% in order of ranking, and weight the analysis results of each model according to the weight to get the final score;

[0021] The scores in the national regional target database are sorted, and the strong, medium, and weak levels are determined according to the quantile of the local area in the national regional target database ranking:

[0022] Strong level: quantile ranks in the top 20% nationwide,

[0023] Middle level: the quantile ranks 20-80% in the country.

[0024] Weak grade: The quantile ranks in the bottom 20% nationwide.

[0025] The present invention also provides an intelligent industrial operation analysis system, including a data processing module, a model management module and an analysis and evaluation module.

[0026] The data processing module processes data based on business requirements and obtains indicator characteristic variables for analysis: it cleans, bins and standardizes public data and government data respectively, and confirms indicator characteristic variables that can be used for model establishment. The indicator characteristic variables confirmed based on government data include: operating performance indicators, development capability indicators, risk resistance indicators, social contribution indicators and green development indicators. The indicator characteristic variables confirmed based on public data include industry scale indicators, business growth indicators, personnel scale indicators and innovation capability indicators.

[0027] The model management module establishes analysis models for government data and public data based on indicator characteristic variables.

[0028] The analysis and evaluation module uses the analysis model of government data to analyze the local regional industrial development situation according to national data labels, and outputs the development level as strong, medium or weak according to the industrial development situation.

[0029] The industry ranking of the local region within the national region is analyzed through an analysis model of public data, and the level is determined according to the percentile of the local region ranking.

[0030] Furthermore, when the data processing module of the intelligent industrial operation analysis system obtains the indicator characteristic variables for analysis, it includes: using the univariate indicator analysis method to identify the characteristic variables with predictive ability for the local regional industry, applying the Pearson correlation coefficient to test the degree of correlation between each related characteristic variable, reducing the collinearity between the characteristic variables through the principal component analysis method, reducing the statistical dimension, and obtaining the final indicator characteristic variable.

[0031] Furthermore, the model management module of the intelligent industrial operation analysis system establishes an analysis model of government data, including: the logical formula for establishing the model is as follows:

[0032]

[0033] w i It is the weight coefficient of each indicator characteristic variable in the overall model f. The initial weight coefficient of the model is preset, and the published digital economy, regional economic indicators, and ranking data are used as the basis for weight system correction to adjust the weight.

[0034] Diff represents the judgment of model f on a specific dimension With country data label y j Be consistent.

[0035] Furthermore, the model management module of the intelligent industrial operation analysis system establishes an analysis model of public data based on the logistic regression algorithm, random forest algorithm, Bagging algorithm, AdaBoost algorithm, voting model algorithm, and neural network algorithm.

[0036] Furthermore, in step 3 of the intelligent industrial operation analysis system, the analysis model of public data is used to define a national target database of regions of the same level according to the local regional level to be assessed as needed.

[0037] For each target in the regional target library, six models are used for analysis;

[0038] Sort the models by their AUC indicators, assign each model a weight of 25%, 20%, 20%, 15%, 10%, and 10% in order of ranking, and weight the analysis results of each model according to the weight to get the final score;

[0039] The scores in the national regional target database are sorted, and the strong, medium, and weak levels are determined according to the quantile of the local area in the national regional target database ranking:

[0040] Strong level: quantile ranks in the top 20% nationwide,

[0041] Middle level: the quantile ranks 20-80% in the country.

[0042] Weak grade: The quantile ranks in the bottom 20% nationwide.

[0043] The benefits of the present invention are as follows: the present invention processes and governs public market data and government data, and through statistical analyses such as monitoring and analysis, high-quality development evaluation, and regional benchmarking analysis, introduces dimensional indicators such as operating performance, development potential, risk resistance, social contribution, and green development based on government data, and introduces dimensional indicators such as industry scale, business growth, personnel scale, and innovation capability that characterize the evaluation of the industrial chain based on public enterprise data, thereby realizing the judgment of the regional industrial development status, and the process of monitoring, analyzing, warning, and predicting the industrial development transformation status, so as to timely discover problems in the process of regional industrial development, and provide support for regional industrial development and decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic flow chart of the method of the present invention. DETAILED DESCRIPTION

[0045] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention.

[0046] Example 1

[0047] The present invention also provides an intelligent industrial operation analysis method, comprising:

[0048] Step 1: Perform targeted data processing based on business requirements to obtain indicator characteristic variables for analysis: perform data cleaning, binning and standardization on public data and government data respectively. The method of the present invention can use big data technologies such as Hadoop, Spark, Flink, etc. to achieve data and indicator processing, and use data processing modules to perform data analysis and standardization.

[0049] Confirm the indicator characteristic variables that can be used for model establishment. The indicator characteristic variables confirmed based on government data include: business performance indicators, development capacity indicators, risk resistance indicators, social contribution indicators and green development indicators. The indicator characteristic variables confirmed based on public data include industry scale indicators, business growth indicators, personnel scale indicators and innovation capacity indicators. When obtaining the indicator characteristic variables for analysis in step 1, it includes: using the univariate indicator analysis method to identify the characteristic variables with predictive ability for local regional industries, applying the Pearson correlation coefficient to test the degree of correlation between the relevant characteristic variables, and reducing the collinearity between the characteristic variables through the principal component analysis method, reducing the statistical dimension, and obtaining the final indicator characteristic variables. Table 1 is the indicator characteristic variables of government data.

[0050] Table 1

[0051]

[0052] Table 2 shows the indicator characteristic variables of public data.

[0053] Table 2

[0054]

[0055] Step 2: Establish analysis models for government data and public data based on indicator characteristic variables respectively.

[0056] In step 2, experts can evaluate and score. Experts can include professors from colleges and universities who evaluate regional industrial development; rating companies who use evaluation methods to evaluate regional industries; and experts from venture capital institutions and VC investors who evaluate industrial development.

[0057] When establishing an analysis model for government data, the following logic formulas may be included:

[0058]

[0059] w i It is the weight coefficient of each indicator characteristic variable in the overall model f. The initial weight coefficient of the model is preset, and the published digital economy, regional economic indicators, and ranking data are used as the basis for weight system correction to adjust the weight.

[0060] Diff represents the judgment of model f on a specific dimension With country data label y j Be consistent.

[0061] The final model output results are that the local regional industrial development is strong, medium, or weak.

[0062] In step 2, analysis models for public data are established based on the logistic regression algorithm, random forest algorithm, Bagging algorithm, AdaBoost algorithm, voting model algorithm, and neural network algorithm.

[0063] Step 3: Analyze the local regional industrial development situation through the analysis model of government data according to national data labels, and output the development level as strong, medium or weak according to the industrial development situation.

[0064] The industry ranking of the local region within the national region is analyzed through an analysis model of public data, and the level is determined according to the percentile of the local region ranking.

[0065] In step 3, six analysis models are used to define a national target database of regions of the same level according to the local regional level to be assessed.

[0066] For each target in the regional target library, six models are used for analysis;

[0067] Sort the models by their AUC indicators, assign each model a weight of 25%, 20%, 20%, 15%, 10%, and 10% in order of ranking, and weight the analysis results of each model according to the weight to get the final score;

[0068] The scores in the national regional target database are sorted, and the strong, medium, and weak levels are determined according to the quantile of the local area in the national regional target database ranking:

[0069] Strong level: quantile ranks in the top 20% nationwide,

[0070] Middle level: the quantile ranks 20-80% in the country.

[0071] Weak grade: The quantile ranks in the bottom 20% nationwide.

[0072] The above-mentioned method of the present invention is based on public data combined with government data to realize the display of the development status of nodes in the local regional industrial chain, and to explore the high-value structure of the entire industrial chain; reveal the regional economic strength, growth potential, affluence, green level, scale structure, innovation level, corporate strength and other situations from various regional indicators, combined with regional industrial layout and situation monitoring, to form a penetrating analysis from the macro, meso and micro levels, and form an overall perception of the overall economic operation; horizontal and vertical comparisons are used to judge the level and gap of its own development, and the development results and trends of the industrial city are presented in a multi-faceted, multi-level and multi-dimensional manner, which is conducive to overall planning, scientific planning and precise policy implementation.

[0073] Example 2

[0074] The present invention also provides an intelligent industrial operation analysis system, including a data processing module, a model management module and an analysis and evaluation module.

[0075] The data processing module processes data based on business requirements and obtains indicator characteristic variables for analysis: it cleans, bins and standardizes public data and government data respectively, and confirms indicator characteristic variables that can be used for model establishment. The indicator characteristic variables confirmed based on government data include: operating performance indicators, development capability indicators, risk resistance indicators, social contribution indicators and green development indicators. The indicator characteristic variables confirmed based on public data include industry scale indicators, business growth indicators, personnel scale indicators and innovation capability indicators.

[0076] The model management module establishes analysis models for government data and public data based on indicator characteristic variables.

[0077] The analysis and evaluation module uses the analysis model of government data to analyze the local regional industrial development situation according to national data labels, and outputs the development level as strong, medium or weak according to the industrial development situation.

[0078] The industry ranking of the local region within the national region is analyzed through an analysis model of public data, and the level is determined according to the percentile of the local region ranking.

[0079] As the information interaction and execution process between the modules in the above-mentioned system are based on the same concept as the embodiment of the method of the present invention, the specific contents can be found in the description of the embodiment of the method of the present invention and will not be repeated here.

[0080] Similarly, the system of the present invention processes and governs public market data and government data through statistical analyses such as monitoring and analysis, high-quality development evaluation, and regional benchmarking analysis. It introduces dimensional indicators such as operating performance, development potential, risk resistance, social contribution, and green development based on government data. It introduces dimensional indicators such as industry scale, business growth, personnel scale, and innovation capability that characterize the evaluation of the industrial chain based on public enterprise data. This enables the judgment of the regional industrial development status, and the process of monitoring, analyzing, warning, and predicting the industrial development transformation status, so as to promptly discover problems in the process of regional industrial development and provide support for regional industrial development and decision-making.

[0081] It should be noted that not all steps and modules in the above-mentioned processes and system structures are necessary, and some steps or modules can be ignored according to actual needs. The execution order of each step is not fixed and can be adjusted as needed. The system structure described in the above-mentioned embodiments can be a physical structure or a logical structure, that is, some modules may be implemented by the same physical entity, or some modules may be implemented by multiple physical entities, or some components in multiple independent devices may be implemented together.

[0082] The above-described embodiments are only preferred embodiments for fully illustrating the present invention, and the protection scope of the present invention is not limited thereto. Equivalent substitutions or changes made by those skilled in the art based on the present invention are within the protection scope of the present invention. The protection scope of the present invention shall be subject to the claims.

Claims

1. A method for analyzing intelligent industrial operation, Its characteristics are include: Step 1: Perform targeted data processing based on business requirements to obtain indicator characteristic variables for analysis: perform data cleaning, binning and standardization on public data and government data respectively, and confirm indicator characteristic variables that can be used for model establishment. The indicator characteristic variables confirmed based on government data include: operating performance indicators, development capability indicators, risk resistance indicators, social contribution indicators and green development indicators. The indicator characteristic variables confirmed based on public data include industry scale indicators, operating growth indicators, personnel scale indicators and innovation capability indicators. Step 2: Establish analysis models for government data and public data based on indicator characteristic variables. Step 3: Analyze the local regional industrial development situation through the analysis model of government data according to national data labels, and output the development level as strong, medium or weak according to the industrial development situation. The industry ranking of the local region within the national region is analyzed through an analysis model of public data, and the level is determined according to the percentile of the local region ranking.

2. The intelligent industrial operation analysis method according to claim 1 is characterized in that when obtaining the indicator characteristic variables for analysis in step 1, include: The univariate indicator analysis method is used to identify characteristic variables with predictive power for local regional industries, and the Pearson correlation coefficient is applied to test the degree of correlation between the relevant characteristic variables. The principal component analysis method is used to reduce the collinearity between characteristic variables, reduce the statistical dimension, and obtain the final indicator characteristic variables.

3. According to the intelligent industrial operation analysis method of claim 1, Its characteristics are In step 2, an analysis model of government data is established, including: The logical formula for establishing the model is as follows: w i It is the weight coefficient of each indicator characteristic variable in the overall model f. The initial weight coefficient of the model is preset, and the published digital economy, regional economic indicators, and ranking data are used as the basis for weight system correction to adjust the weight. Diff represents the judgment of model f on a specific dimension With country data label y j Be consistent.

4. According to claim 1, a method for analyzing intelligent industrial operation, Its characteristics are In the step 2, analysis models for public data are established according to the logistic regression algorithm, random forest algorithm, bagging algorithm, AdaBoost algorithm, voting model algorithm, and neural network algorithm.

5. According to claim 1, a method for analyzing intelligent industrial operation, Its characteristics are In step 3, the analysis model of public data is used to define the target database of regions of the same level across the country according to the local regional level that needs to be assessed. For each target in the regional target library, six models are used for analysis; Sort the models by their AUC indicators, assign each model a weight of 25%, 20%, 20%, 15%, 10%, and 10% in order of ranking, and weight the analysis results of each model according to the weight to get the final score; The scores in the national regional target database are sorted, and the strong, medium, and weak levels are determined according to the quantile of the local area in the national regional target database ranking: Strong level: Percentile ranks in the top 20% nationwide, Middle level: the quantile ranks 20-80% in the country. Weak grade: The quantile ranks in the bottom 20% nationwide.

6. An intelligent industrial operation analysis system, Its characteristics are It includes data processing module, model management module and analysis and evaluation module. The data processing module processes data based on business requirements and obtains indicator characteristic variables for analysis: it cleans, bins and standardizes public data and government data respectively, and confirms indicator characteristic variables that can be used for model establishment. The indicator characteristic variables confirmed based on government data include: operating performance indicators, development capability indicators, risk resistance indicators, social contribution indicators and green development indicators. The indicator characteristic variables confirmed based on public data include industry scale indicators, business growth indicators, personnel scale indicators and innovation capability indicators. The model management module establishes analysis models for government data and public data based on indicator characteristic variables. The analysis and evaluation module uses the analysis model of government data to analyze the local regional industrial development situation according to national data labels, and outputs the development level as strong, medium or weak according to the industrial development situation. The industry ranking of the local region within the national region is analyzed through an analysis model of public data, and the level is determined according to the percentile of the local region ranking.

7. An intelligent industrial operation analysis system according to claim 6, Its characteristics are When the data processing module obtains the indicator characteristic variables for analysis, it includes: using the univariate indicator analysis method to identify the characteristic variables with predictive ability for the local regional industry, applying the Pearson correlation coefficient to test the degree of correlation between the relevant characteristic variables, and reducing the collinearity between the characteristic variables through the principal component analysis method, reducing the statistical dimension, and obtaining the final indicator characteristic variables.

8. An intelligent industrial operation analysis system according to claim 6, Its characteristics are The model management module establishes an analysis model for government data, including: The logical formula for establishing the model is as follows: w i It is the weight coefficient of each indicator characteristic variable in the overall model f. The initial weight coefficient of the model is preset, and the published digital economy, regional economic indicators, and ranking data are used as the basis for weight system correction to adjust the weight. Diff represents the judgment of model f on a specific dimension With country data label y j Be consistent.

9. The intelligent industrial operation analysis system according to claim 6, Its characteristics are The model management module establishes analysis models for public data based on the logistic regression algorithm, random forest algorithm, Bagging algorithm, AdaBoost algorithm, voting model algorithm, and neural network algorithm.

10. An intelligent industrial operation analysis system according to claim 6, Its characteristics are In step 3, the analysis model of public data is used to define the target database of regions of the same level across the country according to the local regional level that needs to be assessed. For each target in the regional target library, six models are used for analysis; Sort the models by their AUC indicators, assign each model a weight of 25%, 20%, 20%, 15%, 10%, and 10% in order of ranking, and weight the analysis results of each model according to the weight to get the final score; The scores in the national regional target database are sorted, and the strong, medium, and weak levels are determined according to the quantile of the local area in the national regional target database ranking: Strong level: Percentile ranks in the top 20% nationwide, Middle level: the quantile ranks 20-80% in the country. Weak grade: The quantile ranks in the bottom 20% nationwide.

Images