Scientific and technological innovation space pattern prediction method under innovation output perspective

Abstract

The invention discloses a scientific and technological innovation spatial pattern prediction method under an innovation output perspective, and relates to the technical field of geographic spatial data processing, and the method comprises the steps: carrying out the preprocessing of obtained scientific and technological innovation data, and constructing a scientific and technological innovation index system based on the preprocessed scientific and technological innovation data; indexes in the scientific and technological innovation index system are analyzed, and an analysis result and a random forest algorithm are combined to construct a scientific and invasive space model; and predicting a science and technology innovation space pattern based on the constructed science and technology innovation space model and in combination with a science and technology project list. Through processing scientific and technological innovation data, innovation output levels of different regions, industries or enterprises can be accurately quantified, and a solid data basis is provided for subsequent spatial pattern prediction; the index system constructed based on the processed scientific and technological innovation data comprehensively covers multiple dimensions of innovation output, and a scientific framework is provided for comprehensively evaluating the innovation output.

Description

Technical Field

[0001] The present invention relates to the field of geographic space data processing technology, and in particular to a method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output. Background Art

[0002] With the continuous advancement of globalization, innovation has become a key force in promoting development. Scientific and technological innovation has not only become a core driving force, but is also closely related to the characteristics of urban form, affecting the development and expansion of urban patterns. At present, research on scientific and technological innovation mainly focuses on the following aspects:

[0003] 1. By analyzing patent cooperation and paper cooperation and applying research methods such as gravity models, we can reveal the innovation links between cities or regions, build innovation networks, and deeply explore the mechanisms behind innovation cooperation.

[0004] 2. Use indicators such as patent output and academic papers to measure the spatial distribution of innovation output and study the spatial spillover effects of innovation activities.

[0005] 3. At the macro level, explore the various factors that affect urban innovation output to identify the key factors that promote or hinder innovation development.

[0006] However, most current studies focus on city-level analysis, and do not fully explore innovation activities within cities. In fact, there is a significant competitive situation among scientific and technological innovation activities in different cities, which to a certain extent affects the continuity and synergy of scientific and technological innovation activities in space. Therefore, if we only examine scientific and technological innovation activities from a macro perspective, it is easy to ignore their inherent spatial attribute characteristics and the spatial proximity effect of knowledge spillovers, that is, the effectiveness of innovation activities is often closely related to their geographical location and surrounding environment. In addition, although a few studies have attempted to touch on the spatial evolution characteristics of innovation activities within cities, these studies often stop at describing the phenomenon and fail to deeply explore the deep-seated mechanisms behind the evolution. Moreover, due to the limitations of statistical data, it is difficult to fully reflect the complexity and diversity of innovation activities within cities.

[0007] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0008] In response to the problems in the related technologies, the present invention proposes a method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output, so as to overcome the above-mentioned technical problems existing in the existing related technologies.

[0009] To this end, the specific technical solution adopted by the present invention is as follows:

[0010] A method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output, the method comprising the following steps:

[0011] S1. Preprocess the acquired scientific and technological innovation data, and build a scientific and technological innovation index system based on the preprocessed scientific and technological innovation data;

[0012] S2. Analyze the indicators in the scientific and technological innovation indicator system, and combine the analysis results with the random forest algorithm to construct a scientific and technological innovation space model;

[0013] S3. Based on the constructed scientific and technological innovation space model and integrating the list of scientific and technological projects, predict the spatial pattern of scientific and technological innovation.

[0014] Furthermore, preprocessing the acquired scientific and technological innovation data and constructing a scientific and technological innovation index system based on the preprocessed scientific and technological innovation data includes the following steps:

[0015] S11. Collect scientific and technological innovation data, which includes address information;

[0016] S12. Use coordinate transformation technology to process address information in scientific and technological innovation data to obtain a spatial data set;

[0017] S13, filter out school data and research institution data from the spatial data set, and merge the school data and research institution data to generate a research and development platform data set;

[0018] S14. Based on data analysis requirements, define the structure of the scientific and technological innovation indicator system, which includes primary indicators and secondary indicators;

[0019] S15. Based on the primary and secondary indicators, combined with the spatial data set and the R&D platform data set, a scientific and technological innovation indicator system is constructed.

[0020] Furthermore, the address information in the scientific and technological innovation data is processed using coordinate transformation technology to obtain a spatial data set, including the following steps:

[0021] S121. Use coordinate conversion technology to geocode the address information in the scientific and technological innovation data to obtain the longitude and latitude coordinates corresponding to each address information;

[0022] S122, importing the longitude and latitude coordinates into a geographic information system to create a vector layer;

[0023] S123. Display the location of scientific and technological innovation data on the vector layer, and add corresponding attribute information to each vector point to obtain a spatial data set.

[0024] Furthermore, the first-level indicators include innovation services, innovation carriers and comprehensive supporting facilities, and the second-level indicators include incubators, maker spaces, R&D platforms, schools, technology companies, living facilities and transportation accessibility.

[0025] Furthermore, the indicators in the scientific and technological innovation indicator system are analyzed, and the analysis results are combined with the random forest algorithm to construct a scientific and technological innovation space model, which includes the following steps:

[0026] S21. Analyze the indicators in the scientific and technological innovation indicator system and calculate the kernel density value, functional mixing value and distance value of the preset area;

[0027] S22, establishing a data set based on the kernel density value, the functional mixing value and the distance value, and dividing the data set into a training set and a test set;

[0028] S23, selecting a decision tree as a base learner of the random forest algorithm, and setting the parameters of the random forest;

[0029] S24, training the random forest model using the training set, evaluating the trained random forest model using the test set, and calculating the determination coefficient of the random forest model;

[0030] S25. Optimize the random forest model according to the determination coefficient and construct a scientific and technological innovation space model.

[0031] Further, analyzing the indicators in the scientific and technological innovation indicator system and calculating the kernel density value, functional mixing value and distance value of the preset area include the following steps:

[0032] S211. Analyze the indicators within the science and technology innovation indicator system, use the kernel density tool to perform spatial calculations on science and technology innovation data, and obtain the kernel density value of the preset area;

[0033] S212, based on the supporting facilities data in the preset area, using the functional mixing degree formula to calculate and obtain the functional mixing degree value of the preset area;

[0034] S213. Based on the traffic accessibility in the preset area, a distance formula is used to measure the distance value from the preset area to the road.

[0035] Furthermore, the indicators in the science and technology innovation index system are analyzed, and the kernel density tool is used to perform spatial calculations on the science and technology innovation data. The kernel density value of the preset area is obtained, which includes the following steps:

[0036] S2111. Analyze the indicators within the science and technology innovation indicator system and collect science and technology innovation data within the preset area;

[0037] S2112. Input the scientific and technological innovation data in the preset area into the geographic information system and set the kernel density analysis parameters of the kernel density tool;

[0038] S2113. Use the kernel density tool that has been set up to perform kernel density analysis on scientific and technological innovation data and generate a kernel density layer;

[0039] S2114. Based on the kernel density layer, obtain the kernel density value of the preset area.

[0040] Furthermore, the functional mixing degree formula is:

[0041]

[0042] In the formula, H parcel Indicates the mixed degree of land use functions of the plot;

[0043] n represents the total number of types of points of interest within the plot;

[0044] i represents the i-th type of interest point;

[0045] P i Indicates the ratio of the i-th type of interest points to the total number of interest points in the plot.

[0046] Furthermore, the distance formula is:

[0047]

[0048] Where D ij Represents the Euclidean distance between point i and point j;

[0049] X i represents the horizontal coordinate of point i;

[0050] Y i represents the ordinate of point i;

[0051] X j represents the horizontal coordinate of point j;

[0052] Y j Represents the ordinate of point j.

[0053] Furthermore, based on the constructed science and technology innovation space model and integrating the science and technology project list, the prediction of the science and technology innovation space pattern includes the following steps:

[0054] S31. Collect a list of scientific and technological projects;

[0055] S32. Match the list of science and technology projects with the science and technology innovation indicator system to obtain new indicators;

[0056] S33, input the new indicators into the science and technology innovation space model, and output the prediction results of the science and technology innovation space;

[0057] S34. Use the spatial connection tools in the geographic information system to combine the spatial prediction results of scientific and technological innovation with geographic spatial data to obtain the predicted spatial pattern of scientific and technological innovation.

[0058] The beneficial effects of the present invention are:

[0059] 1. Through the processing of scientific and technological innovation data, the present invention can accurately quantify the innovation output level of different regions, industries or enterprises, and provide a solid data foundation for the subsequent spatial pattern prediction; the indicator system constructed based on the processed scientific and technological innovation data comprehensively covers multiple dimensions of innovation output, and provides a scientific framework for the comprehensive evaluation of innovation output.

[0060] 2. The present invention significantly improves the prediction accuracy of the scientific and technological innovation space model by applying the random forest algorithm to the analysis results of the scientific and technological innovation index system. It not only enhances the model's ability to capture innovation trends, but also gives the scientific and technological innovation space model greater robustness, providing a reliable basis for decision-making on scientific and technological innovation.

[0061] 3. Based on the scientific and technological innovation space model and combined with the list of scientific and technological projects, the present invention can accurately predict the spatial distribution trend of scientific and technological innovation activities, clarify the innovation potential and advantages of different regions or fields, and thus promote the effective use of innovation resources and the rapid transformation of scientific and technological achievements. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0063] Figure 1 is a flow chart of a method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output according to an embodiment of the present invention;

[0064] Figure 2 is a workflow diagram of a method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output according to an embodiment of the present invention;

[0065] Figure 3 It is a scientific and technological innovation pattern prediction map according to a method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output according to an embodiment of the present invention. DETAILED DESCRIPTION

[0066] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments. They can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and advantages of the present invention.

[0067] According to an embodiment of the present invention, a method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output is provided.

[0068] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1-Figure 3 As shown, according to the method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output in an embodiment of the present invention, the method comprises the following steps:

[0069] S1. Preprocess the acquired scientific and technological innovation data, and build a scientific and technological innovation indicator system based on the preprocessed scientific and technological innovation data.

[0070] It should be noted that the scientific and technological innovation data includes patent data, the list of first-level incubators (list of national incubators), the list of second-level incubators (list of municipal incubators), the list of maker spaces and city POI (points of interest) data.

[0071] Specifically, preprocessing the acquired scientific and technological innovation data and constructing a scientific and technological innovation index system based on the preprocessed scientific and technological innovation data includes the following steps:

[0072] S11. Collect scientific and technological innovation data, which includes address information.

[0073] S12. Use coordinate transformation technology to process the address information in the scientific and technological innovation data to obtain a spatial data set.

[0074] Specifically, using coordinate transformation technology to process the address information in the scientific and technological innovation data to obtain the spatial data set includes the following steps:

[0075] S121. Use coordinate conversion technology to geocode the address information in the scientific and technological innovation data to obtain the longitude and latitude coordinates corresponding to each address information;

[0076] S122, importing the longitude and latitude coordinates into a geographic information system to create a vector layer;

[0077] S123. Display the location of scientific and technological innovation data on the vector layer, and add corresponding attribute information to each vector point to obtain a spatial data set.

[0078] It should be noted that the patent data, incubator list and maker space list all contain address information. The address information is geocoded using coordinate conversion technology (such as Baidu Geocoder API), and the longitude and latitude information corresponding to each address information is exported. Based on the longitude and latitude, the points are displayed in the geographic information system (Arcgis10.6) to obtain patent vector data with attributes such as the main applicant, application date, publication date, publication number and legal status, as well as the vector data of incubators and maker spaces, to obtain a spatial data set.

[0079] S13. Filter out school data and research institution data from the spatial data set, and merge the school data and research institution data to generate the R&D platform data set.

[0080] It should be noted that among the urban POI data, firstly, the school (or university) data are selected based on the category of higher education. Since each college and institution within a university also has separate POI data, in order to avoid affecting the experimental results, such data need to be eliminated, and only one POI data is retained for the same campus of each university. Secondly, based on the name, the data of key laboratories, university science and technology parks, research institutes and other research institutions are selected in the category of scientific research institutions, and they are merged as R&D platform data.

[0081] S14. Based on data analysis requirements, define the structure of the scientific and technological innovation index system. The structure of the scientific and technological innovation index system includes primary indicators and secondary indicators.

[0082] Specifically, the first-level indicators include innovation services, innovation carriers and comprehensive supporting facilities; the second-level indicators include incubators, maker spaces, R&D platforms, schools, technology companies, living facilities and transportation accessibility.

[0083] S15. Based on the primary and secondary indicators, combined with the spatial data set and the R&D platform data set, a scientific and technological innovation indicator system is constructed.

[0084] It should be added that cities are developed and formed by the continuous aggregation and diffusion of economic activities. Against the background of innovation-driven development, urban development has shown a new pattern, and changes in the urban innovation pattern are driven by a variety of factors.

[0085] With the impact of informatization on cities, time and space compression has brought new vitality to the industry, but it has also brought challenges to the traditional economic location theory. Although there is also a spatial proximity effect in innovation activities, this proximity effect is no longer reflected in production materials and other aspects, but is closely centered on talents. Talents are the core of scientific and technological innovation and the main carrier of innovation activities. Attracting talents has become an important task for scientific and technological innovation and urban development. Urban comfort has a strong explanatory power for the spatial flow and agglomeration mechanism of innovative talents. At different urban levels, this comfort may be reflected in macro perspectives such as climate, but within the city, comfort is related to living function facilities and traffic accessibility. Due to the initiative of talents, talents will be more willing to choose areas with convenient living environments, so the comprehensive supporting facilities of the city are also taken into consideration.

[0086] At the same time, the main academic research on scientific and technological innovation is based on Western research, which advocates that the market and capital play a fundamental role in generating innovation activities. In a specific context, a more complex causal relationship is generated between urban development and innovation activities. It can be said that the innovation space is a nurturing innovation space of "active planning", while the innovation space in the West is more like a spontaneous innovation space of "passive planning".

[0087] The market and economic levers are playing an increasingly important role. The innovation environment based on the market itself has energy that cannot be ignored. The innovation environment includes both universities representing knowledge innovation and high-tech enterprises representing industrial innovation. It attaches importance to the changes in the innovation pattern brought about by knowledge spillover and industrial innovation.

[0088] As shown in Table 1, the science and technology innovation index system integrates "innovation services-innovation carriers-comprehensive supporting facilities". Innovation services include incubators, makerspaces and R&D platforms, innovation carriers select universities and high-tech enterprises, and comprehensive supporting facilities include basic living facilities and traffic accessibility. The purpose is to explore the core factors that affect the innovation pattern and provide suggestions and directions for investing in innovation resources. The numerical calculation formula for the first-level category is:

[0089]

[0090] Where, X i represents the value of the i-th first-level category; n represents the number of second-level categories contained in the i-th first-level category; X ij Represents the value of the jth secondary class in the ith primary class.

[0091] Table 1 Science and technology innovation indicator system

[0092]

[0093] S2. Analyze the indicators within the science and technology innovation index system, and combine the analysis results with the random forest algorithm to construct a science and technology innovation space model.

[0094] Specifically, analyzing the indicators in the scientific and technological innovation indicator system and combining the analysis results with the random forest algorithm to construct a scientific and technological innovation space model includes the following steps:

[0095] S21. Analyze the indicators within the science and technology innovation index system, and calculate the kernel density value, functional mixing value and distance value of the preset area.

[0096] Specifically, analyzing the indicators in the scientific and technological innovation indicator system and calculating the kernel density value, functional mixing value and distance value of the preset area include the following steps:

[0097] S211. Analyze the indicators within the science and technology innovation index system, use the kernel density tool to perform spatial calculations on the science and technology innovation data, and obtain the kernel density value of the preset area.

[0098] Specifically, the indicators in the science and technology innovation index system are analyzed, and the kernel density tool is used to perform spatial calculations on the science and technology innovation data to obtain the kernel density value of the preset area, including the following steps:

[0099] S2111. Analyze the indicators within the science and technology innovation indicator system and collect science and technology innovation data within the preset area;

[0100] S2112. Input the scientific and technological innovation data in the preset area into the geographic information system and set the kernel density analysis parameters of the kernel density tool;

[0101] S2113. Use the kernel density tool that has been set up to perform kernel density analysis on scientific and technological innovation data and generate a kernel density layer;

[0102] S2114. Based on the kernel density layer, obtain the kernel density value of the preset area.

[0103] It should be noted that the kernel density analysis method of point elements is used to analyze the evolution of the spatial distribution of innovation outputs, measure the density of innovation outputs in the patent output point and its surrounding area, and measure the degree of agglomeration of innovation outputs, which is used to express the pattern of scientific and technological innovation. At the same time, kernel density analysis is also used to measure indicators such as incubators, maker spaces, R&D platforms, universities, and high-tech enterprises in the scientific and technological innovation index system.

[0104] The kernel density tool in ArcGIS 10.6 is used to calculate the kernel density. In the kernel density analysis, each point is covered with a smooth surface. The surface value is the highest at the point, and the surface value gradually decreases as the distance from the point increases. The surface value is zero at the position where the distance from the point is equal to the search radius. The density of each output grid pixel is the sum of the values ​​of all kernel surfaces superimposed on the center of the grid pixel. The kernel function is based on the quartic kernel function. The calculation formula of the kernel density is as follows:

[0105]

[0106] dist i <radius(i=1,…,n)

[0107] In the formula, Density represents the kernel density value; radius represents the default search radius; n represents the number of points within the search radius; i represents the input point; popi represents the number of patents; dist i Represents the distance between point i and the (x, y) position.

[0108] If the input point is within the radius distance of the point (x, y) position, only the point is included in the sum, and since a single piece of patent data represents a patent record, popi is 1.

[0109] S212: Based on the supporting facilities data in the preset area, the functional mixing degree formula is used to calculate and obtain the functional mixing degree value of the preset area.

[0110] Specifically, the functional mixing formula is:

[0111]

[0112] In the formula, H parcel Indicates the mixed degree of land use functions of the plot;

[0113] n represents the total number of types of points of interest within the plot;

[0114] i represents the i-th type of interest point;

[0115] P i Indicates the ratio of the i-th type of interest points to the total number of interest points in the plot.

[0116] It should be noted that functional mix refers to the complexity of land functions within a certain range. The basic living facilities in the science and technology innovation index system are measured by using functional mix, and the basic living facilities that can be enjoyed within one kilometer of each plot are counted, including seven types of basic living services: housing, education, medical care, entertainment, transportation, financial services and other services. The higher the functional mix, the more living services can be enjoyed within one kilometer of the plot. The urban POI (point of interest) data is divided into seven types: housing, education, medical care, entertainment, transportation, financial services and other services, and the land use functional mix can be calculated.

[0117] S213. Based on the traffic accessibility in the preset area, a distance formula is used to measure the distance value from the preset area to the road.

[0118] Specifically, the distance formula is:

[0119]

[0120] Where D ij Represents the Euclidean distance between point i and point j;

[0121] X i represents the horizontal coordinate of point i;

[0122] Y i represents the ordinate of point i;

[0123] X j represents the horizontal coordinate of point j;

[0124] Y j Represents the ordinate of point j.

[0125] It should be noted that distance analysis is used to measure the traffic accessibility in the science and technology innovation index system, and to measure the distance from the preset area (plot) to the road (urban main road). The Euclidean distance tool in ArcGIS 10.6 is used to perform distance analysis and calculation using the distance formula.

[0126] S22. Based on the kernel density value, the functional mixing value and the distance value, a data set is established, and the data set is divided into a training set and a test set.

[0127] S23. Select a decision tree as the base learner of the random forest algorithm and set the parameters of the random forest.

[0128] S24. Use the training set to train the random forest model, and use the test set to evaluate the trained random forest model, and calculate the determination coefficient of the random forest model.

[0129] S25. Optimize the random forest model according to the determination coefficient and construct a scientific and technological innovation space model.

[0130] It should be noted that, with the block scale as the unit and the complex internal spatial mechanism of the city, traditional quantitative statistical methods are difficult to meet the research needs, so the random forest algorithm is used in Matlab software to explore the factors affecting innovation output. Random forest is a packaged learner that uses the bagging method to integrate decision trees. It votes or averages the prediction results of the base learner decision tree to obtain the final integrated learning result; and "random attribute selection" is added to the training process of the decision tree. The principle of the random forest algorithm is to use multiple decision trees for training and prediction respectively, and use voting or averaging to output the final classification result after obtaining the classification result. The training effect of the model is measured using the determination coefficient (R 2 ) is calculated, the formula is:

[0131]

[0132] In the formula, R 2 represents the determination coefficient; n represents the total number of training samples; α represents the input training sample; z(u α ) represents the true value; R 2 represents the predicted value; R 2 Represents the mean of the predicted values.

[0133] The secondary class as a whole (X ij ), three types in the first-level category (X i) and the individual indicators in the secondary category (X ij ) were input into the random forest toolbox, and the fitting results of each type and indicator are shown in Table 2. The model accuracy is high when the overall input is 2 =0.98), indicating the effectiveness of the model, and the results show that innovation carriers, innovation services and comprehensive supporting facilities are highly correlated with innovation output (R 2 创新服务 =0.76; R 2 创新载体 =0.95; R 2 综合配套 =0.83), among which the correlation of innovation carriers is higher than that of comprehensive supporting facilities and innovation services. At the same time, among the secondary categories, high-tech enterprises have the highest correlation (R 2 高新技术企业 =0.92), followed by colleges and basic living facilities (R 2 高校 =R 2 基本生活配套 =0.86), followed by traffic accessibility (R 2 交通可达性 =0.81), and the correlation of all indicators within the innovative service type did not exceed 0.6 (R 2 孵化器 =0.56; R 2 众创空间 =0.53; R 2 研发平台 =0.60).

[0134] Since innovation output mainly comes from innovation carriers, innovation carriers have become the most important factor affecting the spatial pattern of innovation output. Among innovation carriers, high-tech enterprises carry more innovation output than universities, reflecting that the innovation subject has changed from universities to high-tech enterprises, which is consistent with the conclusion of the transformation of innovation centers. Comprehensive supporting capacity is an important indicator for attracting scientific and technological talents, making comprehensive supporting the second largest factor affecting the pattern of scientific and technological innovation. In comprehensive supporting facilities, basic living facilities and traffic accessibility have a high correlation with the spatial pattern of innovation output. Good basic living facilities can provide more comfortable urban living experience for scientific and technological innovation talents, while convenient transportation can promote the formation and development of scientific and technological innovation space, which is conducive to the gathering of scientific and technological talents and the development of scientific and technological enterprises. Since the impact of innovation services on the spatial pattern of scientific and technological innovation is long-term, the correlation between innovation services and the spatial pattern of scientific and technological innovation is the weakest. Innovation services play a role in promoting scientific and technological innovation incubation, strengthening the combination of scientific and technological innovation and industry, and promoting the adjustment of the scientific and technological innovation pattern. This kind of role is subtle. In innovation services, as R&D platforms with innovation output capabilities, the correlation is higher. Due to the non-short-term nature and poor spatial proximity effect, incubators that only provide innovation incubation and support have a weak correlation with mass innovation spaces.

[0135] Table 2 Training accuracy results of each type and indicator

[0136]

[0137] S3. Based on the constructed scientific and technological innovation space model and integrating the list of scientific and technological projects, predict the spatial pattern of scientific and technological innovation.

[0138] Specifically, based on the constructed science and technology innovation space model and integrating the science and technology project list, predicting the spatial pattern of science and technology innovation includes the following steps:

[0139] S31. Collect a list of scientific and technological projects;

[0140] S32. Match the list of science and technology projects with the science and technology innovation indicator system to obtain new indicators;

[0141] S33, input the new indicators into the science and technology innovation space model, and output the prediction results of the science and technology innovation space;

[0142] S34. Use the spatial connection tools in the geographic information system to combine the spatial prediction results of scientific and technological innovation with geographic spatial data to obtain the predicted spatial pattern of scientific and technological innovation.

[0143] It should be noted that the secondary class indicators are input into the random forest algorithm as a whole, and the learner accuracy (R 2 ) is 0.98, and its prediction results are credible and effective. The obtained scientific and technological innovation space mechanism model can be used to predict the future scientific and technological innovation pattern. According to a list of scientific and technological projects, the scientific and technological innovation center projects and modern service industry projects correspond to the R&D platform (secondary category) in the scientific and technological innovation index system (Table 1), and the advanced manufacturing industry projects correspond to high-tech enterprises (secondary category). Referring to the measurement method in Table 1, the index value (X ij '), X ij 'As the input column, input the science and technology innovation space model, and get the prediction column Y ij ', the prediction column is the innovation output predicted by the model after the implementation of major engineering projects. Using the spatial connection tool in Arcmap, the prediction column is connected with the corresponding plots according to FID to obtain the predicted future scientific and technological innovation space pattern.

[0144] In summary, with the help of the above technical solutions of the present invention, by processing the scientific and technological innovation data, the innovation output level of different regions, industries or enterprises can be accurately quantified, providing a solid data foundation for the subsequent spatial pattern prediction; the index system constructed based on the processed scientific and technological innovation data comprehensively covers multiple dimensions of innovation output, providing a scientific framework for the comprehensive evaluation of innovation output. By applying the random forest algorithm to the analysis results of the scientific and technological innovation index system, the prediction accuracy of the scientific and technological innovation space model is significantly improved, which not only enhances the model's ability to capture innovation trends, but also gives the scientific and technological innovation space model stronger robustness, providing a reliable basis for decision-making in scientific and technological innovation. Based on the scientific and technological innovation space model and combined with the list of scientific and technological projects, the spatial distribution trend of scientific and technological innovation activities can be accurately predicted, and the innovation potential and advantages of different regions or fields can be clarified, thereby promoting the effective use of innovation resources and the rapid transformation of scientific and technological achievements.

[0145] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output, characterized in that: The method comprises the following steps: S1. Preprocess the acquired scientific and technological innovation data, and build a scientific and technological innovation index system based on the preprocessed scientific and technological innovation data; S2. Analyze the indicators in the scientific and technological innovation indicator system, and combine the analysis results with the random forest algorithm to construct a scientific and technological innovation space model; S3. Based on the constructed scientific and technological innovation space model and integrating the list of scientific and technological projects, predict the spatial pattern of scientific and technological innovation.

2. The method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output according to claim 1 is characterized in that: The preprocessing of the acquired scientific and technological innovation data and the construction of a scientific and technological innovation index system based on the preprocessed scientific and technological innovation data include the following steps: S11. Collecting technological innovation data, wherein the technological innovation data includes address information; S12. Use coordinate transformation technology to process address information in scientific and technological innovation data to obtain a spatial data set; S13, filter out school data and research institution data from the spatial data set, and merge the school data and research institution data to generate a research and development platform data set; S14. Define the structure of the scientific and technological innovation indicator system according to data analysis requirements, wherein the structure of the scientific and technological innovation indicator system includes primary indicators and secondary indicators; S15. Based on the primary and secondary indicators, combined with the spatial data set and the R&D platform data set, a scientific and technological innovation indicator system is constructed.

3. The method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output according to claim 2 is characterized in that: The method of using coordinate transformation technology to process the address information in the scientific and technological innovation data to obtain a spatial data set includes the following steps: S121. Use coordinate conversion technology to geocode the address information in the scientific and technological innovation data to obtain the longitude and latitude coordinates corresponding to each address information; S122, importing the longitude and latitude coordinates into a geographic information system to create a vector layer; S123. Display the location of scientific and technological innovation data on the vector layer, and add corresponding attribute information to each vector point to obtain a spatial data set.

4. The method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output according to claim 2 is characterized in that: The first-level indicators include innovation services, innovation carriers and comprehensive supporting facilities; the second-level indicators include incubators, maker spaces, R&D platforms, schools, technology companies, living facilities and transportation accessibility.

5. The method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output according to claim 1 is characterized in that: Analyzing the indicators in the scientific and technological innovation indicator system and combining the analysis results with the random forest algorithm to construct a scientific and technological innovation space model includes the following steps: S21. Analyze the indicators in the scientific and technological innovation indicator system and calculate the kernel density value, functional mixing value and distance value of the preset area; S22, establishing a data set based on the kernel density value, the functional mixing value and the distance value, and dividing the data set into a training set and a test set; S23, selecting a decision tree as a base learner of the random forest algorithm, and setting the parameters of the random forest; S24, training the random forest model using the training set, evaluating the trained random forest model using the test set, and calculating the determination coefficient of the random forest model; S25. Optimize the random forest model according to the determination coefficient and construct a scientific and technological innovation space model.

6. The method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output according to claim 5 is characterized in that: The analysis of the indicators in the scientific and technological innovation indicator system and the calculation of the kernel density value, the functional mixing value and the distance value of the preset area include the following steps: S211. Analyze the indicators within the science and technology innovation indicator system, use the kernel density tool to perform spatial calculations on science and technology innovation data, and obtain the kernel density value of the preset area; S212, based on the supporting facilities data in the preset area, using the functional mixing degree formula to calculate and obtain the functional mixing degree value of the preset area; S213. Based on the traffic accessibility in the preset area, a distance formula is used to measure the distance value from the preset area to the road.

7. The method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output according to claim 6 is characterized in that: The analysis of the indicators in the scientific and technological innovation indicator system and the use of the kernel density tool to perform spatial calculations on the scientific and technological innovation data to obtain the kernel density value of the preset area include the following steps: S2111. Analyze the indicators within the science and technology innovation indicator system and collect science and technology innovation data within the preset area; S2112. Input the scientific and technological innovation data in the preset area into the geographic information system and set the kernel density analysis parameters of the kernel density tool; S2113. Use the kernel density tool that has been set up to perform kernel density analysis on scientific and technological innovation data and generate a kernel density layer; S2114. Based on the kernel density layer, obtain the kernel density value of the preset area.

8. The method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output according to claim 6 is characterized in that: The functional mixing degree formula is: In the formula, H parcel Indicates the mixed degree of land use functions of the plot; n represents the total number of types of points of interest within the plot; i represents the i-th type of interest point; P i Indicates the ratio of the i-th type of interest points to the total number of interest points in the plot.

9. The method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output according to claim 6 is characterized in that: The distance formula is: Where D ij Represents the Euclidean distance between point i and point j; X i represents the horizontal coordinate of point i; Y i represents the ordinate of point i; X j represents the horizontal coordinate of point j; Y j Represents the ordinate of point j.

10. The method for predicting the spatial pattern of scientific and technological innovation from the perspective of innovation output according to claim 1 is characterized in that: The method of predicting the spatial pattern of scientific and technological innovation based on the constructed scientific and technological innovation space model and integrating the scientific and technological project list includes the following steps: S31. Collect a list of scientific and technological projects; S32. Match the list of science and technology projects with the science and technology innovation indicator system to obtain new indicators; S33, input the new indicators into the science and technology innovation space model, and output the prediction results of the science and technology innovation space; S34. Use the spatial connection tools in the geographic information system to combine the spatial prediction results of scientific and technological innovation with geographic spatial data to obtain the predicted spatial pattern of scientific and technological innovation.

Images