The Inﬂuence of Innovation Resources in Higher Education Institutions on the Development of Sci-Tech Parks’ Enterprises in the Urban Innovative Districts at the Stage of Urbanization Transformation

: Innovation is an inevitable way to enhance regional competitiveness, promote urbanization and achieve sustainable development. The sci-tech park is one of the main land use types in the construction of a new town and the critical space carriers of urbanization. This study focuses on the correlation between higher education institutions (HEIs) innovation and sci-tech enterprises (STEs) development, and the Hangzhou West Hi-tech Corridor as a case study. Innovation indicators of HEIs related to teaching sta ﬀ , talent cultivation, scientiﬁc research quality, academic inﬂuence, and university-enterprise cooperation, as well as development indicators of STEs related to aggregation degree, development level, and innovation level are selected. The results show that HEI innovation is signiﬁcantly positively correlated with STE development. The number of senior titled teachers, average number of citations per paper, and number of invested enterprises exhibit a strong positive correlation with STE development indicators. Regarding the correlation between HEI innovation and STE development, academic research HEIs and professional application HEIs are stronger than vocational skill HEIs. This study is conducive to improving land use e ﬃ ciency and a sustainable development level of urban innovative districts. Suggestions are provided to support the researches and practices of urban innovative spatial planning and advancement of the urbanization process.


Introduction
Over the past few decades, urbanization has swept the world and become one of the most remarkable phenomena in the process of global modernization [1]. As the largest developing country in the world, China has experienced long-term and sustained ultra-high speed economic growth and large-scale urbanization since the implementation of reform and opening up in 1978 [2][3][4]. Chinese cities host many important innovative companies and their international connectivities are significantly growing in the world economy [5,6]. From 1989 to 2019, the urbanization rate rose from 26.21% to 60.60% [7], the newly added urban population was about 550 million, and the newly added urban built-up area was more than four times of the original urban area [8]. In other words, more than 80% of area of China's urban has been built in the past 30 years. Among them, the construction of various Land 2020, 9, 396 4 of 36 analysis from the perspectives of industrial structure, functional system, organizational connection, and evaluation [42][43][44]. There are also many successful cases including 22@ Barcelona (Spain), Boston (USA), Macquarie Park (Australia), One-North (Singapore), and Toronto (Canada) innovation districts [41].
The world's first sci-tech park intended to facilitate academic and industrial cooperation can be traced back to 1951 when Stanford University established the Stanford Industrial Park, thus creating a new model of university-industry cooperation [45]. In the 1980s, following the proposal of Professor Frederick Terman concerning the combination of academia and industry, Stanford Industrial Park grew into Silicon Valley [46]. He argued that a research university can contribute to the economic growth of a region, and the development of a region will inevitably lead to the greater development of a university [47]. This marked the emergence of a new economic phenomenon, namely, that of sci-tech parks, which closely linked HEIs and sci-tech industries [48]. Since then, many countries have conducted research and practice on the cooperative relationship between universities and enterprises. The importance of HEIs as external sources of knowledge regarding enterprise innovation has been widely recognized within academic circles.
Most of the researches on the correlation between HEIs and STEs mainly focuses on industr-university-research cooperation. During the early period, many studies were conducted on the related national conditions, government functions, policies, and legal environments. Davies (1983) believed that the proposal of relevant laws would promote the industry-university-research cooperation [49]. Curien (1989) analyzed the role of EUREKA in the promotion of industry-university-research cooperation and considered this bill played a strong catalyzing role in combining the R&D efforts carried out by both industry and universities in Europe [50].
Since the middle to late 1990s, the research on this topic has become more detailed, and breakthroughs have been made regarding research theories and methods. In terms of theoretical research, many studies have been carried out that examine industry-university-research cooperation from different perspectives, such as the perspectives of national innovation theory, triple helix theory, synergy theory, transaction cost theory, resource dependence theory and game theory. Etzkowitz (2003) used triple helix theory to analyze the influence of the relationship between the government, schools and enterprises on collaborative innovation, and argued that they not only maintain their own functions, but also cooperate closely and coordinate with each other to jointly promote the production, transformation, application and upgrading of knowledge in the process of innovation [51]. Mier Zhang et al. (2001) studied the transaction costs of industry-university-research cooperation and believed that these costs have an important impact on the innovation produced by industry-university-research cooperation [52]. In terms of empirical research, many studies focused on the cooperation entities, cooperation motivation, cooperation mode, management mechanisms, evaluation systems, and influencing factors of industry-university-research cooperation, as well as other aspects of this topic. Annama Ria Inzelt (2004) studied the evolution of the government-industry-university-research relationship in the context of Hungary's transition process and found out that Hungarian business create research tasks mainly in experimental development and design, in trials and in the tooling-up process [53]. The Organization for Economic Cooperation and Development (1999) divided the models of industry-university-research cooperation into several categories based on cooperation mode and degree, including those corresponding to general research support, informal cooperative research, contractual research, knowledge transfer and training programs, participation in government joint research programs, R&D alliances, and joint research centers [54]. Gray (2000) examined methodological issues relating to the assessment of university-industry collaboration and concluded interim and outcome assessments have received much evaluative scrutiny since 1980, and the modified peer-review technique has most use [55].
Regarding the impact of universities on enterprise innovation, some studies have aimed to explore the factors that influence the formation of cooperative relations between universities and enterprises. Bojun Hou, et al. (2019) discovered that the efficiency of university-industry collaboration is determined Land 2020, 9,396 5 of 36 by a university's characteristics, the scientific research funds that it receives from the government, and its regional economic status [56]. Xi Li (2020) explored the impact of university-enterprise cooperation on innovation performance in terms of various factors, and his classified review addresses three main perspectives: cooperation network structure perspective, spatial geography perspective and social factors perspective [57]. Maietta, O. W. (2015) examined the drivers of university-firm R&D collaboration and assessed the determinants of innovation in a low-tech industry [58]. Some scholars generally believe that geographic proximity between universities and companies promotes connections between these organizations. Hervas Oliver, Albors Garrigos (2009) proposed that geographic proximity between HEIs and enterprises is very important for innovation, as it is conducive to the establishment of connections between them and to transfers of knowledge, especially that of tacit knowledge [59]. D'Este et al. (2007) argued that geographical proximity plays a fundamental role in university-industry cooperation [60]. Garcia, Renato et al. (2013) suggested geographic proximity between academic research and firms' research and development (R&D) facilities matters in fostering university-industry linkages [61]. Maietta, O. W. (2015) studied the driving factors behind school-enterprise cooperative innovation and believed that enterprises within 150 km of a university have more product innovation possibilities than those located far from a university [58]. The importance of geographic proximity to technological dynamism and economic competitiveness has been questioned as the development of science and technology has brought about a reduction in the cost of transportation and an improvement in the speed and quality of long-distance interactions. However, Anselin L et al. (2000) believed that the obscurity of knowledge determines that it must be acquired through direct and interpersonal communication [62]. He also argued that as long as there is a time interval between the generation and coding of knowledge, the primary mechanism of knowledge flow is the interaction between people (2004) [63]. Enterprises can acquire knowledge faster and more successfully through local and person-to-person contact.
HEIs are an important source of knowledge, technology, and other innovation resources (Wei Xu, 2003) [64]. The existing studies on this topic have proposed a number of indicators that reflect the development of HEIs and that of enterprises . Florian Findler et al. (2019) analyzed to what extent sustainability assessment tools are capable of measuring the impacts that HEIs have on Sustainable Development [65]. Donglin Yang and Bo Meng (2010) noted that the index of scientific research funds has become an important index by which to evaluate the comprehensive abilities of universities [66]. Patent data reflect the effective innovation output of organizations, and various studies have verified the reliability of the distribution characteristics and development of innovation activities based on patent data [67]. David (2015) used US patent and citation data to measure the technical relevance between major patent categories and to link them to various knowledge networks [68]. Tartari and Breschi (2012) noted that there is a positive correlation between the share of a researcher's academic achievements that are published in applied journals and his or her propensity to work with industry partners [69]. Van Looy et al. (2011) argued that the size of a university, measured by the number of academics within it, affects the extent to which that university produces innovation and turns "outwards" [70]. Fontana, R et al. (2006) proposed that the staff size of an enterprise affects its university-industry cooperation and that larger enterprises are often more likely to benefit from scientific research [71]. According to the 2019 "Scoring Standard Rules of National Sci-tech Enterprises Identification", intellectual properties is an important indicator for the measurement of the innovative ability of enterprises [72].
In conclusion, since the establishment of Stanford Industrial Park in the 1950s, most of the studies on the correlation between higher education institutions and technological innovation ability have focused on industry-university-research cooperation. It is generally believed in the academic circles that the industry-university-research cooperation is conducive to the development of all cooperative entities and the enhancement of the comprehensive strength of urban innovation districts. The theoretical research mainly includes the formulation of cooperation related policies and regulations, the proposal and application of cooperation theoretical model, cooperation model, driving force and influencing Land 2020, 9, 396 6 of 36 factors. In practice, excellent cases like Silicon Valley, Zhongguancun and 22@Barcelona have also appeared.
The following gaps can be found based on previous studies. First, most of the existing works have carried out empirical studies on the cooperation modes, management mechanisms, and influencing factors of industry-university-research cooperation from the perspective of a single cooperative partner in an enterprise, university, or research institution. However, there are few studies focused on the correlation among various cooperative partners. Second, some studies have proposed a number of indicators for the measurement of universities' and enterprises' development; however, there is a lack of systematic research on the relationship between these two entities in the context of innovation, especially the comparative study of the correlation between different types of HEIs and STEs. Third, academics generally believe that geographical proximity promotes cooperation and connection between universities and enterprises. However, the studies in this field are mostly of a qualitative nature and lack quantitative analyses. Therefore, this study attempts to narrow the above gaps on the basis of previous studies and from the perspective of innovation-driven social and economic sustainable development. It quantitatively analyzes the correlation between the two entities of industry-university-research cooperation; namely, the innovation resources of HEIs and the development of the STEs close to these HEIs, and attempt to identify the innovation indicators related to the influence of HEIs on the development of STEs.
HEIs and STEs. Third, academics generally believe that geographical proximity promotes cooperation and connection between universities and enterprises. However, the studies in this field are mostly of a qualitative nature and lack quantitative analyses. Therefore, this study attempts to narrow the above gaps on the basis of previous studies and from the perspective of innovation-driven social and economic sustainable development. It quantitatively analyzes the correlation between the two entities of industry-university-research cooperation; namely, the innovation resources of HEIs and the development of the STEs close to these HEIs, and attempt to identify the innovation indicators related to the influence of HEIs on the development of STEs.

Study Area
The Corridor is located in western Hangzhou, spanning east to west across 15 streets within three districts; namely, Xihu District, Yuhang District, and Lin'an District. With Wenyi West Road as its main axis, the Corridor's boundaries stretch from the Zijingang Campus of Zhejiang University in the east to the Zhejiang Agriculture and Forestry University in the west, passing through the Zijingang Science and Technology City, the Future Science and Technology City, and the Qingshanhu Science and Technology City (see Figure 1). It is approximately 33 km long and has a planned total area of approximately 224 square kilometers. According to the "Spatial Master Plan of the Hangzhou West Hi-tech Corridor" and the current land use status of the area, the Corridor is divided into 31 sections [73]. These areas are the central area, the scientific research area, the higher education area, the industrial area, the residential area, the comprehensive area, and the leisure area (see Figure 2). The industries in the Corridor is focused on future networks, big data cloud computing, e-commerce, artificial intelligence, life sciences, new energy vehicles, new materials, technology services, and new finance. It is the forefront of China's "Internet Plus" and "Artificial Intelligence Plus" industries; additionally, it is the main venue for innovation in global technology and industry.

Study Area
The Corridor is located in western Hangzhou, spanning east to west across 15 streets within three districts; namely, Xihu District, Yuhang District, and Lin'an District. With Wenyi West Road as its main axis, the Corridor's boundaries stretch from the Zijingang Campus of Zhejiang University in the east to the Zhejiang Agriculture and Forestry University in the west, passing through the Zijingang Science and Technology City, the Future Science and Technology City, and the Qingshanhu Science and Technology City (see Figure 1). It is approximately 33 km long and has a planned total area of approximately 224 square kilometers. According to the "Spatial Master Plan of the Hangzhou West Hi-tech Corridor" and the current land use status of the area, the Corridor is divided into 31 sections [73]. These areas are the central area, the scientific research area, the higher education area, the industrial area, the residential area, the comprehensive area, and the leisure area (see Figure 2). The industries in the Corridor is focused on future networks, big data cloud computing, e-commerce, artificial intelligence, life sciences, new energy vehicles, new materials, technology services, and new finance. It is the forefront of China's "Internet Plus" and "Artificial Intelligence Plus" industries; additionally, it is the main venue for innovation in global technology and industry.

Data Regarding Development of Sci-Tech Enterprises
This study collects sci-tech enterprises information in the Corridor from Qichacha snd Tianyancha, which are the official query platforms of National Enterprise Credit Information Publicity System. The data collection process steps are as follows: Step 1: Determine the collection scope. Using the map query function of Qichacha, draw several circles on the map to ensure that these circles can fully cover the scope of the Corridor.
Step 2: Eliminate duplicate enterprises' data. Collect all kinds of information such as enterprise name, coordinates, industry type, staff size, the number of intellectual properties, the number of patents, and the content of sci-tech innovation in each circle on Qichacha. Summarize the information and eliminate duplicate enterprises, and then generate enterprise information summary in Excel.

Data Regarding Development of Sci-Tech Enterprises
This study collects sci-tech enterprises information in the Corridor from Qichacha snd Tianyancha, which are the official query platforms of National Enterprise Credit Information Publicity System. The data collection process steps are as follows: Step 1: Determine the collection scope. Using the map query function of Qichacha, draw several circles on the map to ensure that these circles can fully cover the scope of the Corridor.
Step 2: Eliminate duplicate enterprises' data. Collect all kinds of information such as enterprise name, coordinates, industry type, staff size, the number of intellectual properties, the number of patents, and the content of sci-tech innovation in each circle on Qichacha. Summarize the information and eliminate duplicate enterprises, and then generate enterprise information summary in Excel.
Step 3: Eliminate the out-of-scope enterprises' data. According to the coordinates, the information of each enterprise is imported into ArcGIS (geographic information system platform software). Each enterprise can be seen as a point element. According to the boundary line of the Corridor, eliminate the enterprises beyond the boundary range.
Step 4: Eliminate non-science and technology enterprises: according to the sci-tech and innovation industry type determined in Hangzhou West Hi-tech Corridor Planning [74], the enterprises not engaged in the sci-tech and innovation industry will be eliminated.
Step 5: Supplement enterprises data: According to the name of the enterprises, collect comprehensive score data of STEs on Tianyancha.
Finally, the information of 16,504 sci-tech enterprises in the Corridor is collected. This study takes the development indicators of STEs as its dependent variables and includes the number of enterprises surrounding HEIs (Y1), which reflects the degree of STE aggregation; the comprehensive scores (Y2) and staff sizes (Y3) of STEs, which are used to reflect their levels of development; the number of intellectual properties (Y4), the number of patents (Y5) and the content of sci-tech innovation (Y6) owned by STEs, which are used to reflect their levels of innovation.
This study views the centroid of the spatial outline of each higher education institution as the center of a circle with a radius of 3 km; this circle represents the area to be sampled (see Figure 3). Since some of the universities and colleges are located on the borders of the Corridor, the sampled areas stretch beyond the borders of the Corridor. The relationship between the HEIs and the STEs surrounding them cannot be accurately reflected if only the enterprises inside the Corridor are accounted for in the statistics. Therefore, this study first calculates the area where each circle intersects with the Corridor; then, the indicators value of all the enterprises in a circle is sum up separately; finally, the ratio of the sum of the indicators value to intersection area to be used as a statistic describing the enterprises.
Land 2020, 9, x FOR PEER REVIEW 8 of 38 Step 3: Eliminate the out-of-scope enterprises' data. According to the coordinates, the information of each enterprise is imported into ArcGIS (geographic information system platform software). Each enterprise can be seen as a point element. According to the boundary line of the Corridor, eliminate the enterprises beyond the boundary range.
Step 4: Eliminate non-science and technology enterprises: according to the sci-tech and innovation industry type determined in Hangzhou West Hi-tech Corridor Planning [74], the enterprises not engaged in the sci-tech and innovation industry will be eliminated.
Step 5: Supplement enterprises data: According to the name of the enterprises, collect comprehensive score data of STEs on Tianyancha.
Finally, the information of 16,504 sci-tech enterprises in the Corridor is collected. This study takes the development indicators of STEs as its dependent variables and includes the number of enterprises surrounding HEIs (Y1), which reflects the degree of STE aggregation; the comprehensive scores (Y2) and staff sizes (Y3) of STEs, which are used to reflect their levels of development; the number of intellectual properties (Y4), the number of patents (Y5) and the content of sci-tech innovation (Y6) owned by STEs, which are used to reflect their levels of innovation.
This study views the centroid of the spatial outline of each higher education institution as the center of a circle with a radius of 3 km; this circle represents the area to be sampled (see Figure 3). Since some of the universities and colleges are located on the borders of the Corridor, the sampled areas stretch beyond the borders of the Corridor. The relationship between the HEIs and the STEs surrounding them cannot be accurately reflected if only the enterprises inside the Corridor are accounted for in the statistics. Therefore, this study first calculates the area where each circle intersects with the Corridor; then, the indicators value of all the enterprises in a circle is sum up separately; finally, the ratio of the sum of the indicators value to intersection area to be used as a statistic describing the enterprises.

Data Regarding Innovation of Higher Education Institutions
• Geographic Spatial Data The geographic spatial data includes HEIs data and STEs data. Among them, the geographic spatial data of HEIs mainly refers to the geographical location and the land use of each HEI in the Corridor. The data comes from the land use status map of Hangzhou city, the land use of HEIs in the land use map of Hangzhou West Hi-tech Corridor Planning, and the geographic spatial scope in the planning of each HEI. The data is mainly used to define the spatial distribution of HEIs in the

Data Regarding Innovation of Higher Education Institutions
• Geographic Spatial Data The geographic spatial data includes HEIs data and STEs data. Among them, the geographic spatial data of HEIs mainly refers to the geographical location and the land use of each HEI in the Corridor. The data comes from the land use status map of Hangzhou city, the land use of HEIs in the land use map of Hangzhou West Hi-tech Corridor Planning, and the geographic spatial scope in the planning of each HEI. The data is mainly used to define the spatial distribution of HEIs in the Corridor and to determine the sampling range by locating the centroid position of each HEI. The geographic spatial data of STEs mainly refers to the distribution of each enterprise in the Corridor. According to the coordinates of each enterprise, enterprise data can be imported into ArcGIS. Through geographic Land 2020, 9, 396 9 of 36 spatial data, the relationship between the distribution of STEs and the location of HEIs can be observed, which also provides space conditions for further analyzing the influence of innovation resources in higher education institutions on the development of surrounding sci-tech parks' enterprises.
• Innovation Data The data sources employed include the "Compilation of Science and Technology Statistics of Higher Education Institutions [73]", which was compiled by the Department of Science and Technology of the Ministry of Education; the "Annual Report on the Employment Quality of Graduates [75]", which was published in 2019; the "Annual Report on the Quality of Higher Vocational Education [76]", which was published by the Employment Guidance Centers of HEIs; an institutional analysis of the Wanfang Data Knowledge Service Platform; and the official website of each higher education institution examined.
In 2019, the Ministry of Education of the People's Republic of China launched a survey to assess the innovation of HEIs, including that of undergraduate institutions, junior colleges, and independent colleges. The survey covered innovative talent cultivation, teaching staff, industry-university-research cooperation, and innovative technology achievement transformation. Based on this survey and the previous studies, this study selects 17 indicators to reflect the innovation of HEIs that correspond to five categories: teaching staff, talent cultivation, quality of scientific research, academic influence, and university-enterprise cooperation. This study takes these 17 indicators as its independent variables; namely, the number of teaching staff (X1), the number of teachers with senior titles (X2), the number of teaching and research staff (X3), the number of R&D staff (X4), the number of graduates (X5), the number of graduates engaged in sci-tech innovation industry (X6), the number of intellectual properties (X7), the number of patents (X8), the number of sci-tech achievements (X9), the number of sci-tech topics addressed (X10), the number of articles issued by journals (X11), the content of sci-tech innovation (X12), the funds for scientific research (X13), the number of cited papers (X14), the average number of citations per paper (X15), the number of invested enterprises (X16), and the number of holding enterprises (X17).

Spatial Analysis
ArcGIS is used to analyze the spatial distribution of the enterprises and HEIs, as well as the kernel density of the development indicators of the STEs in the Corridor. The teaching staff, talent cultivation, quality of scientific research, academic influence, and university-enterprise cooperation of the HEIs are identified and visualized.

Partial Least Squares Regression (PLSR)
PLSR was first proposed by S. Wold and C. Albano in 1983. This method consists of a combination of principal component analysis, canonical correlation analysis and multiple linear regression analysis [77]. First, the principal component analysis principle is applied to condense the information contained in the multiple independent variables X and dependent variables Y analyzed and obtain the principal component of these variables (X corresponds to the principal component U, Y corresponds to the principal component V). Then, the relationship between X and U, and the relationship between Y and V are analyzed by means of the typical correlation principle. Finally, the relationship between X and V is analyzed via multiple linear regression to identify the relationship between X and Y. PLSR is a multivariate statistical method that can solve collinearity problems, simultaneously analyze multiple dependent variables Y, and deal with relationships in small samples.
Supposing that the p dependent variables Y1, . . . , Yp and the m independent variables X1, . . . , Xm are all standardized variables, the n-times standardized observation data matrix of the dependent variable group and that of the independent variable group are denoted as follows, respectively: Step 1: Determine the optimal number of principal components through cross-validity analysis and variable importance in projection (VIP) analysis. For the cross-validity analysis, if Qh2 ≤ 0.0975, this means that it is not beneficial to continue increasing the number of principal components; that is, the number of principal components being used at this point is the optimal number to utilize. For the VIP analysis, if there is not an obvious change in the VIP value when the number of principal component is increased, the number of principal components currently being utilized is optimal.

•
Step 2: Extract the principal components and perform an accuracy analysis. The principal component U1 is extracted from the independent variable X, and the principal component V1 is extracted from the dependent variable Y. The components U1 and V1 must extract as much information as possible from the variable groups that they are in; thus, the degree of the correlation between U1 and V1 will be maximized. An accuracy analysis examines the information extraction rate (variance interpretation rate) of the principal components of X and Y and analyzes the model effect.

•
Step 3: Conduct a PLSR analysis. Analyze the correlation between X and Y as identified by the regression, including the regression coefficient, the significance of the relationship, and the R square value.
This study focuses on the correlation between HEIs and the STEs surrounding them. Due to the small sample size and the high level of collinearity between the variables, a PLSR method is used for this study.

Spatial Distribution of Sci-Tech Enterprises
According to the statistics regarding the number of enterprises in different areas, the Future Science and Technology City Central Area, the Jiangcun Central Area, the West Lake Sci-tech Innovative Parks Scientific Research Area, the Hangzhou University City Higher Education Area, the Xiaoheshan Higher Education Area, and the Zhejiang University Xixi Higher Education Area contain the most STEs (see Figure 4).
According to a kernel density analysis of the distribution of the enterprises, it can be observed that the STEs examined are mainly located in the Zijingang Science and Technology City and the Future Science and Technology City, which are closer to downtown Hangzhou. The Qingshanhu Science and Technology City Area has fewer STEs than the abovementioned areas, and only a small number of STEs are located in the Qingshanhu Science and Technology City central area and the Cloud Manufacturing Township scientific research area (see Figure 5).

Analysis of the Development of the Sci-Tech Enterprises
In terms of development scale, the enterprises with larger staff sizes and higher comprehensive scores are concentrated in the Future Science and Technology City and the Zijingang Science and Technology City areas in the eastern part of the Corridor, and are especially concentrated in the Future Science and Technology City Central Area, the West Lake Science and Technology Park Scientific Research Area, the Jiangcun Central Area, and the Zhejiang University Xixi Higher Education Area.
In terms of innovation level, the Future Science and Technology City Central Area still has the highest innovation level; this area has the most intellectual properties and patents, and the highest level of sci-tech innovation content. This area contains a large number of sci-tech parks and enterprises. In addition, the Hangzhou University City Higher Education Area, the Taobao City Comprehensive Area, the West Lake Sci-tech Innovative Parks Scientific Research Area and the Wuchang Comprehensive Area along the principal axis of Wenyi West Road exhibit relatively high levels of innovative ability. The innovative ability of the Qingshanhu Science and Technology City in the western region is relatively low.

Classification and Distribution of Higher Education Institutions
According to the "International Standard Classification of Education (ISCED) [78]" issued by UNESCO in 2011, and the "Suggestions on the Establishment of Higher Education Institutions [79]" issued by the Ministry of Education of China in 2017, and due to the current situations in our country regarding the entrance of higher education into its popularization stage and the current social talent structure, the HEIs in the Corridor are classified into three types: • Type I-Academic research: This type of HEI focuses on discovering theories regarding basic disciplines and applied disciplines and cultivates leading innovative academic talent.

•
Type II-Professional application: This type of HEI focuses on acquiring high-tech professional knowledge pertaining to various industries and the ways in which high-level technology can be transformed into productivity; this type of HEI cultivates applied talent at different levels.

•
Type III-Vocational skills: Most of these HEIs are vocational colleges and mainly train professional and technical talent for future engagement in production, management, service and other specific types of work.
According to the quantitative statistics that describe the area, there are 22 HEIs in the Corridor, and these HEIs have a wealth of higher education resources (see Table 1). However, there are relatively few academic research colleges and universities, namely, only Zhejiang University and Westlake University. Additionally, with the exception of Hupan University, there is a lack of high-end vocational and technical colleges for the cultivation of high-end vocationally skilled workers. In terms of distribution, most of these colleges and universities are located on the boundaries of the Corridor. Figure 5 shows that the distribution exhibits characteristics that could be described as "scattered on the border and far away from the hinterland". Only a small number of colleges and universities are located in the Xiaoheshan Higher Education Area (see Figure 6).

Innovation Indicators of Higher Education Institutions
This study measures the innovation ability of the examined HEIs with 17 indicators corresponding to five categories: teaching staff, talent cultivation, quality of scientific research, academic influence, and university-enterprise cooperation (see Table 2). There are 3 academic research universities, and all of them are located in the Zijingang Science and Technology City. These three universities are two campus of Zhejiang University, the only "985 university (global first-class high-level university)" in Zhejiang Province, and Westlake University in the Yungu Scientific Research Area. There are 9 professional application universities and colleges, and these include provincial undergraduate colleges and independent colleges. Among these schools, Hangzhou Dianzi University Information Engineering School, Zhejiang Agriculture and Forestry University and Hangzhou Medical College are scattered in Qingshanhu Science and Technology City. Hangzhou Normal University and Zhejiang University of Science and Technology are located in the Hangzhou University City Higher Education Area. The remaining professional application universities and colleges are concentrated in the Xiaoheshan Higher Education Area. There are 10 vocational skill universities and colleges, and these are mainly vocational colleges and technical schools. These schools are scattered; however, they are generally located in the Future Science and Technology City and the Zijingang Science and Technology City.

Innovation Indicators of Higher Education Institutions
This study measures the innovation ability of the examined HEIs with 17 indicators corresponding to five categories: teaching staff, talent cultivation, quality of scientific research, academic influence, and university-enterprise cooperation (see Table 2).
The two campuses of Zhejiang University are located in the eastern part of the Corridor. As it is the only C9 (the first top universities alliance in China) university in Zhejiang Province, the value of each indicator for Zhejiang University is far higher than the values of the indicators corresponding to the other universities and colleges; thus, its innovative ability is higher. In addition, the HEIs examined are somewhat concentrated in the Xiaoheshan Higher Education Area and the Hangzhou University City Higher Education Area, so the innovative ability of this area is relatively high.
In terms of teaching staff, the values of the academic research and professional application indicators are significantly higher than those of the other categories, especially in the case of Zhejiang University and the Zhejiang University of Technology. These schools have obvious advantages in terms of the number of teachers with senior titles that they employ and their R&D staff. In terms of talent cultivation, academic research and professional application schools mainly focus on producing applied and innovative talent, and the number of graduates from these schools that work in the sci-tech innovation industry is larger than that of other schools. Vocational skill schools focus on producing technical talent. In terms of quality of the scientific research generated and the academic influence of the institutions, both academic research and professional application schools have an absolute advantage in terms of the number of their sci-tech achievements, the content of their sci-tech innovation, and the funds that they receive for scientific research. In terms of university-enterprise cooperation, with the exception of Zhejiang University, the Zhejiang University of Technology, and Zhejiang Agriculture and Forestry University, the number of invested enterprises and holding enterprises are very small in other universities and colleges.

Relevance of Location and Site Selection
By overlapping the kernel density of the STEs with the spatial distribution map of the HEIs (see Figure 7), the relevance of location selection can be clearly seen: the 7 areas of the Corridor in which the STEs are concentrated are all geographically close to universities or colleges; additionally, they are located in the areas where the HEIs are clustered. For example, the border between the Xiaoheshan Higher Education Area and the Liuxia Comprehensive Area, where many enterprises are clustered, is adjacent to many colleges in the Xiaoheshan Higher Education Area; the Zijingang Campus of Zhejiang University is surrounded by a group of STEs, and these STEs are especially concentrated in the Westlake Science and Technology Park Research Area and the Jiangcun Central Area.       are located in the areas where the HEIs are clustered. For example, the border between the Xiaoheshan Higher Education Area and the Liuxia Comprehensive Area, where many enterprises are clustered, is adjacent to many colleges in the Xiaoheshan Higher Education Area; the Zijingang Campus of Zhejiang University is surrounded by a group of STEs, and these STEs are especially concentrated in the Westlake Science and Technology Park Research Area and the Jiangcun Central Area.

Correlations among Innovation Indicators of All the Higher Education Institutions and the Development Indicators of the Sci-Tech Enterprises
By observing the distribution of higher education institutions and sci-tech enterprises in the Corridor, it can be seen that the enterprises are basically distributed within three kilometers around the HEIs. The development status of the STEs located near universities and colleges is determined based on the aggregation degree, development scale and innovation ability of each STE. The correlations among innovation indicators of all the HEIs and the development indicators of the STEs would be analyzed by using PLSR.
Step 1: Perform principal component analysis by cross-validity analysis and VIP analysis to confirm that the optimal number of principal components is 1 (see Table 3). The cross-validity can be reflected by the value of Q h 2 . Step 2: principal components U1 and V1are extracted, and accuracy analysis is performed. The comprehensive extraction ratio of principal component U1 for all 17 independent variables is 0.955 (variance interpretation rate, i.e., 95.5%), with a high extraction ratio. Among them, X15 has a high information extraction ratio of 0.602. The information extraction ratio of other 16 independent variables is all higher than 0.9 (see Table 4). The comprehensive extraction ratio of principal component V1 for all 6 dependent variables is 0.939 (variance interpretation rate, i.e., 93.9%), with a high extraction ratio. The principal component V1 has high information extraction ratios for all 6 dependent variables Y, which were 0.959, 0.961, 0.904, 0.963, 0.921 and 0.928 (see Table 5). Step 3: Analyze the correlation between X and Y.
The results of the PLSR analysis are as follows (see Figure 8 and Table 6):    In conclusion, seven HEI innovation indicators, namely, the number of teaching staff, number of teachers with senior titles, number of teaching and research staff, number of graduates engaged in the sci-tech industry, average number of citations per paper, number of invested enterprises, and number of holding enterprises have a significant positive influence on the number of STEs that surround these institutions. Ten HEI innovation indicators, namely, the number of teaching staff, number of teachers with senior titles, number of graduates, number of patents, number of sci-tech topics, number of articles issued by journals, number of cited papers, average number of citations per paper, number of invested enterprises, and number of holding enterprises, have a significant positive influence on the development level of the surrounding STEs. Three HEI innovation indicators, namely, the number of teachers with senior titles, average number of citations per paper, and number of invested enterprises have a significant positive influence on the innovation level of the surrounding STEs. The indicators denoting the number of teachers with senior titles, average number of citations per paper, and number of invested enterprises have the highest correlation with the enterprise development indicators.

Correlations between the Innovation of the Different Types of Higher Education Institutions and the Development of Sci-Tech Enterprises
Six indicators are selected from the above 17 indicators as independent variables, and a PLSR is performed with these six independent variables and the dependent variables to analyze the correlation between the different types of HEIs and the development of STEs. These six independent variables are the number of teachers with senior titles (X1), number of graduates engaged in the sci-tech innovation industry (X2), number of sci-tech achievements (X3), number of articles issued by journals (X4), number of cited papers (X5), and number of invested enterprises (X6). The regression results are as follows (see Tables 7 and 8

Conclusions and Discussion
The future development of sci-tech parks are the major land use types and space carriers of urban development. From the perspective of innovation-driven social and economic development, sci-tech parks and enterprises must employ their existing innovative elements to guarantee the diversified support of sci-tech industries. As they are an important source of innovation factors, such as intelligence, technology, and talent, the various types of HEIs will provide strong support for the development of STEs in the future.

Conclusions and Discussion
The future development of sci-tech parks are the major land use types and space carriers of urban development. From the perspective of innovation-driven social and economic development, sci-tech parks and enterprises must employ their existing innovative elements to guarantee the diversified support of sci-tech industries. As they are an important source of innovation factors, such as intelligence, technology, and talent, the various types of HEIs will provide strong support for the development of STEs in the future.

Conclusions and Discussion
The future development of sci-tech parks are the major land use types and space carriers of urban development. From the perspective of innovation-driven social and economic development, sci-tech parks and enterprises must employ their existing innovative elements to guarantee the diversified support of sci-tech industries. As they are an important source of innovation factors, such as intelligence, technology, and talent, the various types of HEIs will provide strong support for the development of STEs in the future.
This study focuses on the correlation between HEI innovation and STE development. The Hangzhou West Hi-tech Corridor is selected as the case study because it exhibits the latest trend of urbanization development in China. The Corridor is dominated by the sci-tech industry and is an innovative space that is composed of various innovative factors, such as sci-tech parks, HEIs, and scientific research institutes. PLSR is used to analyze the correlations between the teaching staff, talent cultivation, quality of scientific research, academic influence, and university-enterprise cooperation of HEIs, and the aggregation degree, development level and innovation level of STEs. Accordingly, this study on the correlation between the agglomeration of innovation resources in HEIs and development of sci-tech parks' enterprises can optimize the allocation of innovation resources in urban innovative districts and improve the land use efficiency and regional comprehensive innovation level. Futhermore, it provides references for constructions of sci-tech parks in urban innovative districts and industry-university-research cooperation, and supports for the advancement of urbanization and regional innovation and sustainable development.
The main research conclusions of this paper are as follows: (1) Overall, there is a significantly positive correlation between the innovation of HEIs and the development of the surrounding STEs. The indicators denoting the number of teachers with senior titles, average number of citations per paper, and number of invested enterprises of HEIs exhibit the highest correlation with the enterprise development indicators.
As the source of knowledge, technology, intelligence and talents, HEIs have played an important role in the development of urban innovative districts, especially the surrounding STEs. Improving the senior teaching staff, academic influence and the number of invested enterprises in HEIs may promote the development of STEs in the surrounding areas. Therefore, in the urban innovative spatial planning, the spatial layout of HEIs and STEs should be optimized to strengthen the spatial connection between them, and the university-industry cooperation mechanism should be improved in order to form a collaborative innovation circle of industry-university interaction, namely, "branch campus of top academic research and professional application HEIs + high-end vocational skill HEIs + sci-tech parks".
(2) The teaching staff of HEIs has a significant positive influence on the comprehensive scores of enterprises. The academic influence of HEIs has a significant positive influence on the comprehensive scores, staff sizes and number of intellectual properties of STEs. Furthermore, university-enterprise cooperation has a significant positive influence on the number of enterprises surrounding HEIs, as well as the comprehensive scores and staff sizes of STEs.
Therefore, it can be inferred that increasing the number of teachers and improving the quality of high-level teachers of HEIs would be conducive to improving the comprehensive scores of the surrounding enterprises. Enhancing the academic influence of HEIs would be conducive to promoting the development levels and innovation levels of the surrounding STEs. Increasing the number of investing and holding enterprises of HEIs would be conducive to the agglomeration of the surrounding enterprises and to the enhancement of the development levels of the surrounding STEs. between vocational skill HEIs and the development indicators of STEs. The explanatory power of the independent variables to explain the dependent variables in academic research HEIs and professional application HEIs is generally more powerful than that of vocational skill HEIs.
In the future urban innovative districts planning, the existing innovation resources should be anchored, and the innovative leading effects of academic research universities should be highlighted to strengthen the spillover of innovation elements to STEs. The first is to introduce high-quality innovative resources by strengthening cooperation with top universities at home and abroad and encouraging them to settle in and build branch schools. The second is to improve the innovation quality of high-end vocational skill schools to make up for the shortage of senior technical personnel training. The contributions of this study can be seen more clearly when it is compared with several previous studies. Some previous studies have discussed the factors that influence industry-university-research cooperation from different perspectives. Jiuru Zhao (2018) believed that when selecting university partners, enterprises should first evaluate the status, scale, reputation and scientific research quality of the universities, including the number of scientific researchers, the number of patent, and the number of co-authors' papers [80]. Maietta Ornella Wanda (2015) proposed that the factors affecting university-industry cooperation include academic research quality, university scale, faculty composition, and researcher qualifications [58]. Xi Peng & Min Xue (2016) concluded that the research quality of universities is one of the factors influencing the cooperation [81]. The influencing factors mentioned in these studies support some of the results of this study. These innovation resources of HEIs have a positive impact on the development of STEs and can promote the industry-university-research cooperation. However, these studies are mostly based on empirical analysis of cooperation cases or qualitative research summarized by previous experience, which is different from the relevance of quantitative analysis in this study. Also, the previous studies only put forward the influencing factors of industry-university-research, but did not specify whether these factors have a positive or negative impact and what aspects they have affected. In this study, indicators are extracted from the two subjects of HEIs and STEs respectively for regression analysis. The results can reflect the influence of different innovation resources on different aspects of STEs' development.
Some scholars argued that professional application HEIs are more likely to carry out industry-university-research cooperation. Arvanitis (2005) pointed out that professional application HEIs are closer to the technology needs of enterprises and are more likely to collaborate with enterprises [82]. Dichu Xiao (2014) and Haifeng Liu (2014) considered that the most important thing of establishing professional application HEIs is to carry out industry-university-research cooperation, which indicates that professional application HEIs are more suitable for industry-university-research cooperation [83,84]. These results prove that professional application HEIs have a stronger correlation and greater influence on enterprise development than vocational skill HEIs. However, they have neglected the role of academic research HEIs. Compared with the other two kinds of HEIs, the academic research HEIs have a stronger level of innovation resources, which may have a larger radiation capacity and influence scope for the surrounding STEs.
In summary, this study mainly focuses on the sci-tech parks, one of the new land use types, and the correlation among different innovative subjects in the urban innovative districts at the stage of urbanization transformation. This manuscript mainly focuses on the influence of innovation resources in higher education institutions on the development of sci-tech parks' enterprises. Our findings fill the gap in quantitative research on the correlations among different innovative subjects in the context of innovative development and find out the indicators of innovation factors in HEIs which are significantly correlated with the development of STEs.
This study mainly explores the influence of various innovative resources in HEIs on the development of STEs in sci-tech parks around them. As an important base for the transformation of HEIs' scientific research achievements, sci-tech parks are major carriers for the future urban development and a key node in the urban innovation space network. They provide a platform for HEIs and research institutions to transform their research achievements so as to realize the value of research achievements. On the other hand, the input of various innovative resources such as knowledge, technology and talents also promotes the development of the park's sci-tech and innovation industry and the improvement of its innovation level.
The connection between HEIs and sci-tech parks is an important part of urban innovation space network. In the practice of industry-university-research, the innovation influence of HEIs is not only limited to the surrounding STEs, but also radiated to the surrounding cities, regions and even foreign countries. This is particularly evident in academic research HEIs and professional application HEIs. For example, Zhejiang University undertakes the tasks of decision-making consultation, academic research and talent training and has provided scientific and technological services to more than 3000 enterprises and institutions in China. It has also established cooperative relations with universities, research institutes, sci-tech parks and enterprises in Europe, the United States, Canada, Japan, and other regions and countries. At present, the national strategies of regional coordinated development and Yangtze River Delta integration are gradually advancing. As a key core city in the Yangtze River Delta area, Hangzhou's innovative resources have broken provincial boundaries and radiate to neighboring cities such as Shanghai, Suzhou, Hefei, et al. HEIs cooperate with sci-tech parks in surrounding cities to build laboratories, research and development centers, innovation practice bases and other projects. The connection between them has expanded from the inner city to the periphery of the city, and then to the overall regional innovation and collaborative development of the urban agglomeration.
There are some limitations to this study with respect to indicator selection and the areas sampled. First, when selecting the indicators of HEI innovation and STE development, this study referred to previous studies on this topic and the relevant national policies, regulations and investigation reports. However, only some of the available indicators were selected as the analysis variables, which may have had an impact on the research results. Additionally, this study set the areas to be sampled for sci-tech enterprise data to a radius of three kilometers around each HEI due to the actual distribution of the HEIs and STEs in the Corridor, a factor that may have also affected the research results.
Future research will discuss the effects of different innovation activity subjects on the development of sci-tech parks and industry-university-research cooperation. Furthermore, it will also explore the land use changes, space-time evolution and spatial planning in the urban innovative districts in the context of urbanization. The results will be verified through the comparative analysis in different urban innovative districts.