Construction of Student Innovation and Entrepreneurship Experience System Integrating K-Means Clustering Algorithm

e traditional talent development model is unlikely to satisfy the demands of the times and socioeconomic growth. Innovationdriven development policies require a great deal of creativity and entrepreneurship. Student creativity and education in the enterprise have emerged as an essential source of concern for all segments of society. e traditional talent cultivation mode has been unable to meet the requirements of the economic growth of the society. College graduates are national talents of high quality and high level, which are important talent resources for colleges and universities to grow sustainably. Innovation and entrepreneurship education for college students is a common concern of the current society. e research and implementation of innovation and entrepreneurship education for Chinese college students are progressing rapidly, but the innovation and entrepreneurship education of Chinese college students began later and is not at a high level of development at present. Promoting student innovation and entrepreneurship requires further improvements in research and teaching methods. is article presents the scientic classication and the K-means algorithm. It denes the concepts related to the education system of change and entrepreneurship, the concepts related to the search for change and the theories of entrepreneurship, and the theoretical foundations of the development of the education system. Of the researchers, 39.47% were male; 60.53% were female; 13.68% were students; 70.53% were undergraduate students; and 15.79% were postgraduate students.


Introduction
Innovation capacity is an important indicator of a country's competitiveness. Innovation is a key driver of development and a strategic pillar for building a modern economic system. All countries attach great importance to fostering innovative and entrepreneurial capabilities in order to stay ahead of economic globalization and contribute to economic development. University students are the country's highly educated talent and form the intellectual backbone and engine of sustainable development. People are paying more attention to innovation and the entrepreneurial economy, and the era of entrepreneurship and innovation is growing, which undoubtedly requires us to cultivate more pioneering talents who can adapt to entrepreneurship and have entrepreneurial awareness, entrepreneurial quality, and entrepreneurial knowledge.
In the current era of economics and higher education, it is important to encourage students to start their own businesses, adapt innovation and education to entrepreneurship, and guide them to focus on developing entrepreneurial skills in the processes of change and on entrepreneurship education [1]. Stimulating the full development of students is important and appropriate. It has important theoretical and practical implications for China's social and economic development, the reduction of pressure at work, the enrichment of innovation in learning and entrepreneurship, and the personal development of students.
Regarding cluster analysis, relevant scientists have done the following research. Sacha et al. [1] present a multistep visual analysis method for iteratively re ning clustering and performing time series analysis using self-organizing graphs. is method proves to be e ective for interactive data analysis and helps improve the clarity of clustering results and the interactive process itself [2]. Khan et al. [3] used cluster analysis and logistic regression to identify the most important social and environmental factors experienced by children living in New York and to assess whether those who had some type of asthma already had asthma. Children who live in rented apartments, densely populated buildings, and older buildings are more likely to develop asthma than other children [3]. Ma et al. [4] proposed a cluster analysis strategy to determine the average daily heat consumption of university-level buildings. e defined profile also helps classify different blocks into groups based on the same specific energy consumption actions to favor further projects on energy performance [5]. Hammou et al. [5] proposed an anisotropic distribution process to improve the smoothness of concave and convex valleys. Although maintaining boundaries between them, comparative results show results from the proposed group method [6]. Zheng et al. [6] proposed a range of models for pixel-based cluster analysis to develop climate relationships using remote sensing and temperature data. e results of this study will be used to analyze the environmental consequences of climate change and to increase awareness of future environmental changes [7]. Jang and Smyth [7] updated the full-scale finite element model of a large-scale long-span bridge to improve consistency with actual measured data. ey apply two methods to improve the model update process. On the basis of sensitivity analysis, they performed a cluster analysis to find a very efficient set of update parameters [8]. Nayak et al. [8] proposed a hybrid approach of new algorithms for optimizing cluster analysis. ey considered performance metrics such as error rate and inter-and intracluster distances. ey simulated rigorous experiments, which showed that the suggested hybrid approach fared better than other methods [9]. Lindsey et al. [9] recommended using basic and group component analysis to determine methods for comparing test sites with valid sources. A result is a convenient tool for estimating potential research targets with inaccurate or incomplete information [10]. Maghsoodi et al. [10] proposed a new comprehensive approach to simplify decision analysis with hypotheses based on cluster sequences that follow the precise integration of large data structures. e proposed approach is based on a combination of data processing methods that follow company and domain standards [11]. Diakhate et al. [11] compared the evolution of crack length estimation on the basis of experimental analysis (cluster analysis of sound radiation activity) and numerical simulations. ey demonstrated the remarkable power of acoustic radiation technology, as well as statistical tools to control the growth of cracks in the wood [12]. Data analysis methods are often used in stock exchanges and inventory analysis. Tekin and Gümüs [12] minimized human intervention by creating algorithmic processes that make the most appropriate choices and increase revenue. Shares are classified based on financial indicators obtained from the company's financial statements [13]. Neri et al. [13] applied cluster analysis to an index framework that presents the interaction of three dimensions of durability: environmental, social, and economic.
is is a useful and integrated tool to assess the performance of countries in a timely manner and to improve the criteria for ranking countries in terms of sustainable development [14]. Lankton et al. [14] proposed four propositions about possible user policies. ese policies are used in combination with different disclosure levels, network sizes, and privacy settings. Using cluster analysis, they found some support for all four propositions, as well as additional strategies that beginners might employ [15]. Wang et al. [15] used an automated clustering algorithm to find clusters with atomic densities other than the data. is method only requires independent variables; other inputs can be estimated or determined on the basis of physical variables such as network parameters and concentrations solved [16]. is method provides guidelines for our study, but due to the short study duration and small sample size, this study did not receive general approval. e document is innovative in that it "supports the full development of students." A system of innovation and entrepreneurship education for students will be established, which will include a system of goals, an information system, a support and guarantee structure, a training system, and an evaluation system. is article does not address any aspects of innovation and entrepreneurship in higher education. e article proposes an algorithm for grouping K-means and scientific calculations for the analysis and processing of innovation and entrepreneurship data.

K-Means Clustering Algorithm.
Cluster analysis is an important area of data science and a fundamental technique for finding distributions and patterns in data. e clustering problem can be described as classifying a set of data points so those in each cluster are as close as possible to each other and those in separate clusters are as different as they can be. e main advantages of the K-means algorithm for clustering are as follows: (1) it is a classical method for clustering questions that is easy, fast, and simple to implement; (2) for large data sets, there is no need to specify the distance matrix, which is scalable and efficient; and (3) if the final clustering result shows that the data distribution of each subcategory is very dense and the difference between subcategories is large, the clustering result is good. A system architecture layout for the algorithm is shown in Figure 1.
Cluster analysis, as an unsupervised learning method, applies the techniques of multivariate analysis to numerical analysis. e process of reasonably classified large amount of preprocessed data is according to its own characteristics when there is no reference or basis for the dataset. It aggregates different data into clusters according to their own intrinsic properties and relationships with other data objects.
ere is a high degree of similarity between data placed in the same group and little similarity or difference between data placed in different groups. e higher the probability inside the cluster and the difference among clusters, the more effective the data grouping is. e data set can be represented by a matrix; each row represents an object, and each corresponding column is the attribute characteristic value of the object. e realization of the clustering algorithm needs to go through the process of data preprocessing, feature selection and extraction, clustering, and evaluation of effectiveness. In data preprocessing, the final data collected has a lot of missing values, noise, incompleteness, or invalidity, which makes data analysis impossible. e foundation of data analysis is based on data quality. In order to ensure the quality of data, data cleaning is the first step in data analysis. In this process, data consistency, invalidity, and missing values need to be dealt with. e processing of consistency mainly uses the method of checking and checking and correction, and the processing of invalidity and missing values is processed by the method of data cleaning, including estimation and attribute deletion.
In feature selection and extraction, when analyzing objects that need to be clustered, the features and noise contained in actual objects are complex, and the complexity of clustering is related to the number of results and feature values. erefore, in order to reduce the complex data analysis, it is possible to perform dimensional reduction processing on the data points. at is, selecting functions or separating functions selects some of the influencing functions after examining the set of functions to achieve the size reduction goal. Feature extraction is a method of projecting the feature vector of high latitude to the space of low dimensionality using function mapping, thereby reducing the dimension. In practical applications, the best results will be achieved by applying the two methods together.
Appropriate adjacent words are chosen to describe the similarities or differences between the analyzed objects. e similarity between two objects is represented by a numerical value of the degree of similarity between the objects, that is, the larger the numerical value, the more similar the objects are. e similarity is calculated based on the distance between objects; a smaller distance means a greater similarity between two objects, and vice versa.
where S is the absolute distance.
where S(M, N) is the Euclidean distance.
where S(m, N) is the Chebyshev distance.
where a v is the mean of the sample and b v is the number of samples in the classification.
where D * v is the average squared distance between samples between classes.
where a v is the sample mean vector and a is the mean vector of all samples.
where B is the number of data in the data set and Z u is the center of the cluster.
where B v is the number of objects.
where Q u -cluster centroids, b u -the number of data points, and u up -data objects within a cluster.
where cosθ is the cosine distance between eigenvectors.
where l is the class divergence of clustering.
where s 1 , s 2 is the data objects and se is the number of data objects.
where M u , M v are the data objects in the sampled data set.
where b is the number of data objects in the subclass and M v is a data object in a subclass.
Although clustering research has been carried out for a long time, with the widespread use of databases, the popularization of the Internet, and the development of multimedia, clustering applications face many new challenges: (1) validity. Traditional clustering methods can only identify convex and spherical groups, while in reality many groups are convex and complex. Accurate identification of complex clusters remains one of the hottest research topics. As the database grows, the clusters in the database are no longer uniform. at is, there are clusters of different sizes, densities, and shapes in the database, and sometimes, the difference between clusters and groups is not obvious, so how to find these clusters is also a topic that needs to be studied. Almost all group testing is done to ensure validity because that is what group testing is for. (2) Results. With the advent of large-scale databases and the development of the largest databases in the world, the classification of large and complex data has become a major challenge. At the heart of this problem is finding algorithms that effectively measure quantitative and quantitative data. (3) Clustering of multidimensional and multivariate data and complex data. e new application focuses on clustering with databases that do not contain the same type of data. Clusters of objects can be very large and have their own structure. Many clustering algorithms no longer depend on the size or scale of the data. e difference between grouping and comparing is that the required grouping class is unknown. e collection is the process of dividing data into different categories or groups so that targets in the same group are nearly identical, but targets in separate groups are very different.
In performance analysis, no matter what clustering algorithm is used in the database, there are different clusters. But even clusters found using different methods in the same database can be different and require different methods. Generally speaking, they can be classified as either external keys or as internal keys. e internal key is an analysis of the clustering results and the data in the database, usually a comparison function defined by the value of a function in the database. e lower the comparison function's value, the more consistent the classification result is with the task. In the outer key, the performance of the database is evaluated using nondata information by taking the results of the agglomerative clustering algorithm and comparing them to an external database and comparing the requirements of the two databases. e flowchart of the algorithm is shown in Figure 2.

Scientific Computing.
Scientific computing is the process of using machines to reproduce, predict, and discover the laws of motion and evolutionary features of the objective world. Scientific computing is the application of computers to perform numerical calculations in order to solve mathematical problems in science and technology. Scientific computing has many applications, and a large number of computer programs and application software have been developed to model and analyze specific research problems, including computational fluid dynamics.
ere are generally two development directions for the improvement of computing performance. One is to improve computing performance by increasing the clock frequency of a single processor. Another direction is to improve computing performance by increasing the number of computing cores on the processor chip. However, a series of problems such as the power consumption wall brought about by the increase of the core frequency make the improvement of computing performance brought about by the increase of the clock frequency to the end. e power consumption of a single-core processor is roughly proportional to the cube of its main frequency. When the computing performance is improved by increasing the number of cores, the power consumption increases linearly. erefore, multicore platforms have become the mainstream direction to improve computing performance. In order to take advantage of the computing performance of the multicore platform, it is necessary to carry out corresponding parallelization research on scientific computing algorithms.
Traditional scientific computing is done through mainframes or large-scale computer clusters, so the parallelization research of scientific computing algorithms on mainframes or clusters has been relatively mature. However, with advances in processor manufacturing skills and the growth of on-chip multicore architectures, the computing performance of embedded processor systems has been rapidly improved. Meanwhile, the advantages of low energy usage and ease of cutting are becoming increasingly apparent. As a result, increasing attention is being paid to reallife applications in production and life. e application of scientific computing has gradually expanded from laboratory research to real-life applications. For example, image processing in life often requires the use of algorithms from scientific computing, such as matrix calculations. e implementation of parallelization of scientific computing algorithms on an embedded multicore platform is therefore of great practical importance. As shown in Figure 3, it is a classification diagram of scientific computing visualization.
In scientific computing, data visualization can be divided into two broad categories: symbolic and structured data and images and symbols. e basic process of data visualization is to first preprocess the original input data by filtering and use the mapping algorithm to capture the visual data of the application to map the data. It captures geometric data to visualize the geometric data, captures image data, and finally delivers it to the final visualization tool. e architecture parameters of the microprocessor are shown in Table 1.

2.3.
Students' Innovation and Entrepreneurship. Innovation and entrepreneurship are activities related to the creation of social wealth. It gradually increases social wealth through the continuous improvement and development of the entire society's economic development and scientific level. As far as innovation and entrepreneurship are concerned, the essence of this concept refers to the practice of innovation and entrepreneurship through its innovative business behavior, which is generally the activities of people participating in social production. It is a practical process of innovation and creation through social production and practical work. Innovative entrepreneurs can learn about entrepreneurship and innovation, develop innovative and entrepreneurial skills, raise their awareness of innovation and entrepreneurship, promote their growth and development, and engage in creative ventures. e growth of entrepreneurship and innovative awareness must be based on scientific theories shaped by scientific concepts.
Exam-oriented education, quality education, innovation education, and entrepreneurship education have a historical background and a theoretical basis. Training in innovation and entrepreneurship is offered today, not taken for granted. It has a theoretical spirit that must be followed and has its theoretical basis and basis. Specifically, there are three theoretical foundations of Marx's theory of all-round development of human beings, Marxist innovation theory, and subjectivity education theory. e connotation of innovation and entrepreneurship covers multiple levels and aspects. It is a process of constant change and self-improvement, built on breaking down old systems and breaking mindsets. It includes establishing new development mechanisms and ways of thinking and then expanding development space, creating new things, Randomly select k data from n data objects as initial cluster centers Practical innovation and entrepreneurship are integral parts of innovation and entrepreneurship. Innovative and entrepreneurial education is different from traditional teaching. As a new paradigm of education, it is increasingly adapted to the development requirements of the contemporary world. ey improve higher education to improve the quality and promote the general growth of students. However, the characteristics of innovative and entrepreneurial education are embodied in the aspects below. e first is creativity.
rough learning concepts and new teaching methods, teachers develop creativity and entrepreneurship and develop students' understanding of change and creative thinking in an organized and integrated manner. is increases the courage of students to deal with their initial knowledge and how to deal with it. rough education, innovation, and entrepreneurship, students create creative images and use scientific thinking to research and acquire new knowledge and technologies. e second is creation. e value of education in creativity and entrepreneurship lies in respecting the quality of the individual and educating students in the best possible way. ey combine student characteristics and introduce different teaching methods for different types of students. e third is efficiency. Innovation and entrepreneurship education is different from traditional education because they emphasize practical education. ey provide opportunities and platforms for undergraduates to have the opportunity to gain entrepreneurial experience and help students transform their knowledge and talents for the better. Figure 4 shows the content and specific concepts of innovation and entrepreneurship education. It can be seen that education and courses are still an important part of the educational and entrepreneurial innovation that students see. It is believed that innovation and entrepreneurial capacity, awareness, and enthusiasm contribute significantly to innovation and entrepreneurial education.

Experiments and Result Analysis
e e extraction process of the K-cluster algorithm implies as many partition sizes as possible. If the sampling capacity is very small between each class, the K-mean algorithm can generally provide better results. However, if the pattern similarity between species is great, there may be more clusters. erefore, it is possible that due to a combination of algorithms, they get a local rather than an international rating. e smallest solution uses a supervised classification method for the supervised classification of experimental data. e selected classifier is the maximum likelihood classifier, and the classification results obtained are used to verify the accuracy of the clustering results obtained by the dynamic algorithm. e specific flowchart of the algorithm is shown in Figure 5.
After selecting the approximate optimal parameters, replace the variables in the source program with constants whose values are the optimized parameter values obtained by the selection and then obtain the optimized program through local compilation. For simplicity, when the approximate optimal parameters obtained by the test selection correspond to the execution time, these parameter values are still put into the parameter file. As a program without reduction transformation, the input parameters of the executable program obtained by compiling execute it to get the optimized time. As shown in Figure 6, it is a schematic diagram of the flow of scientific and technological algorithms.
e main population of this study was undergraduate students. e basic characteristics of the respondents are shown in Tables 2-4.       Mathematical Problems in Engineering Figure 7 shows the distribution of entrepreneurial stress among people of different backgrounds. It can be seen that 60% of the respondents have a business plan. A closer look shows that men are more likely to start a business than women. University students are more excited about entrepreneurship than students at the undergraduate level, and students are more excited about starting their own businesses than postgraduate students.
A survey of students with diverse entrepreneurial backgrounds found that students with experience in entrepreneurship received lower ratings from teachers than students with no business experience. Many innovators in teaching innovation and entrepreneurship do not have their own work or practical experience in innovation and entrepreneurship, and there are mentors and leaders who also become teachers and do not respond to students' needs. Figure 8 shows the impact of pedagogical innovation and entrepreneurship on teacher education and the impact of entrepreneurial experience on teaching teacher innovation and entrepreneurship at different levels of education. Figure 8 shows that many innovation and entrepreneurship educators do not have creative work, entrepreneurship, or hands-on experience. ere are also teachers and administrators who guide teachers and do not meet the needs of students. Figure 9 provides information about forms of innovation and entrepreneurship education.
is paper describes how students are more likely to receive creative and educational forms in entrepreneurship. e top half of the diagram in Figure 9 is for a suitable creative and entrepreneurial training session. ey are participating in entrepreneurship competitions, participating in entrepreneurship experimental parks, engaging in part-time jobs, self-learning innovation and entrepreneurship knowledge, and others.
Chinese colleges and universities are influenced by the traditional educational philosophy. e traditional educational philosophy believes that academic excellence leads to an official career. Whether it is school education or family education, it is believed that being a civil servant is the best way out after reading. Although China's development has given college students other options than civil servants, the employment goal of most college students is to find a stable job. ey choose career psychology because they require stability and low risk, do not accept entrepreneurial behavior, and even characterize entrepreneurial behavior as a helpless act, thinking that college students entrepreneurs choose to start a business because they cannot find an ideal job. Although some college students envy college students who are successful in entrepreneurship, they still have no confidence in choosing to start a business. When many college students start their own business, what they get is not encouragement and support, but a skeptical attitude.
Entrepreneurship is a high-IQ activity that requires advanced skills and strong abilities to use knowledge, transform it into external form, and then turn it into actual productivity. In this process, the key skills that entrepreneurs need are social networking skills, practical skills, and leadership skills. erefore, students need to continuously cultivate these three abilities before and during the entrepreneurial process. rough ideological and political education, students' ideological motivation changes, and their internal motivation is enhanced, encouraging them to consciously cultivate their entrepreneurial ability. at is to say, through ideological education and civic education, it conveys the truth that "if people do not advance, they will retreat." It awakens their anxiety, enhances their learning motivation, and mobilizes their enthusiasm, initiative, and creativity in learning. It makes them realize that only by working hard, actively developing themselves, improving their entrepreneurial ability, and expanding their abilities in all aspects can they be invincible in the entrepreneurial competition.
e promotion of entrepreneurship education should focus on cultivating the psychological and educational quality of students' entrepreneurial work and enhancing self-confidence, courage, emotions, and other healthy personalities. Entrepreneurial quality is the psychological quality of entrepreneurship, and it is the characteristic that guides entrepreneurial psychology and entrepreneurial attitude. It is closely related to human instinct and nature and reflects the feelings and emotions of entrepreneurs. e better psychological quality includes the psychological quality of independent and rational thinking, the psychological characteristics of entrepreneurial flexibility, the psychological quality of entrepreneurial risk management, and the entrepreneurial quality psychology of entrepreneurial preparation. And personality is not only a psychology course but an important ideological and political education course. Ideological and political education not only can create conditions for learning and improve the mental qualities needed for entrepreneurship but also can strengthen the quality of entrepreneurship. us, the main task of universities and colleges is to grow and direct students to continuously strengthen their personalities and entrepreneurship and develop creative personalities that meet the needs of society. It is also an important topic in education, business, ideas, and politics. is does not mean Mathematical Problems in Engineering that students will face many difficulties and obstacles in the entrepreneurial process. Without good mental qualities and personality, it is impossible to start a business. us, the entrepreneurial education process should aim to improve the mental quality of entrepreneurs, strengthen the education and courage of healthy individuals such as self-confidence, and improve the quality and personality of entrepreneurs. e development of innovation and entrepreneurship education curriculum is shown in Figure 10.

Discussion
e natural attributes of human beings profoundly reveal the most essential physiological and material needs of human beings. Its development is influenced by the social relations in which it is located, and its social attributes can only be satisfied when it develops to a certain level of society. And it is necessary to realize the development of one's own innovation and entrepreneurship ability, and the development of ability first comes from the germination of consciousness.
However, as a single individual, it is restricted by its own subjective factors and influenced by objective environmental factors, and it gradually forms different social classes. As a result, each individual's ability to innovate and start a business is different. It further affects people's spiritual attributes, causing everyone to have various differences and inconsistencies in cognition, behavior, and abilities, and then affects the entire society. e most fundamental way to solve these dilemmas is to cultivate people's awareness of innovation and entrepreneurship, enhance innovation and entrepreneurship ability, coordinate and adjust social relations, and develop socialist productive forces. Only in this way can the harmonious development of human's natural attributes and social and spiritual attributes be realized.
In the final analysis, the cultivation of innovation and entrepreneurship awareness is the manifestation of people's awareness of problems, questions, criticism, and innovation. It is the manifestation of the psychological motivation of all subjective activities of the subject, which is the spirit of criticism and questioning that is constantly produced. People's subjective thinking is constantly mobilized and stimulated, prompting them to put forward new ideas for solving difficulties. Based on independent thinking, they fully mobilize their own autonomy, initiative, and innovation. e essence of its subjectivity is highly sublimated. Innovation is the transcendence and breakthrough of people's existing knowledge and practice. It is the highest manifestation of human subjectivity and also the deepseated basis for the creative effect of human beings. It is the display of human subjectivity that brings into play the diversity, independence, and innovation of human personality, and the cultivation of innovation and entrepreneurship awareness fully reflects the sublimation of human subjectivity development.
Only by doing a good job of teaching talents can we bring the potential of undergraduate students into high play and help them become successful talents. Different undergraduate students have different attitudes towards innovation and entrepreneurship, different interests in innovation and entrepreneurship, and different levels of innovation and entrepreneurship. Different educated people have different characteristics. Innovation and entrepreneurship education must take a gradual and focused approach to solving the practical problems of the educational process in order to    provide guidance and guidance in a more targeted and effective way.

Conclusion
Under the new normal of the economy, the pressure on social and economic activities has increased; reforms in some areas have continued to deepen; and the overall economic scale has remained stable. Innovation and entrepreneurship are new sources of entrepreneurial vitality for economic development. is will contribute to a new phase of economic development, open a new phase of industrial reform, and accelerate the transformation and restructuring of the existing economy. is paper states that education, innovation, and entrepreneurship for students not only emphasizes the teaching of theoretical knowledge of innovation and entrepreneurship but also emphasizes the development of practical innovation and entrepreneurial skills in students and in global high schools. e introduction of the K-means clustering scientific computing algorithm provides a more scientific direction for student innovation and entrepreneurship.
is will improve the quality of innovation and entrepreneurial skills in general and enhance the competitiveness of university students in the business world. is thesis makes a preliminary forecast of this, but due to the source of data and the limitations of academic standards, some omissions are inevitable in this thesis. e analysis in the status quo analysis stage is incomplete, showing only the changes in the indicators in question and lacking the analysis to make internal judgments about them; in theory, it has not been fully mastered in theory. Due to limited time, the relevant materials collected are not comprehensive enough. Whether these materials can objectively and truly reflect the current situation still needs to be verified through innovation and entrepreneurship education research and practice.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.