Research on the Construction of Industrial College and Talent Cultivation Mode under the Perspective of Industry-Education Integration--Taking “Seven-Color Craftsmen” Auto Repair Industrial College as an Example

In recent years, in order to cultivate high-quality applied, compound and innovative skilled talents, higher vocational colleges and universities have gradually deepened the integration of industry and education and school-enterprise cooperation, and industrial colleges have emerged. The construction of industrial colleges is an effective measure for higher vocational colleges and universities to take the initiative to adapt to the new industry and new trends under the new situation of industry-education integration, to integrate resources and to promote the market-oriented operation of higher vocational education resources [1–4].

As a new type of educational platform with multiple subjects, industry college is a community of destiny with close integration of schools and enterprises and related interests, which is an essential improvement to the general loose school-enterprise cooperation mode from the perspectives of governance concepts, organizational structure and operation mechanism, and has a very good role to play in solving the deep-rooted problems of school-enterprise cooperation, such as low enthusiasm for participation of enterprises, poor stability of cooperation, poor quality of teaching and difficult to promote the integration of production and education, thus better serving the local economy and regional industries [5–8]. It has a very good promotion effect to solve the deep-rooted problems of school-enterprise cooperation, such as poor stability of cooperation, unsatisfactory teaching quality, and difficulty in promoting the integration of production and education, so as to better serve the local economy and regional industries [5–8]. At the same time, because the industrial college is run by the government, enterprises and schools of multiple educational subjects, its diversified education system significantly improves the depth and breadth of the integration of resources in all aspects, and also greatly enhances the mutual integration of various educational subjects, promotes the fit between various resources and various subjects, and greatly promotes the in-depth integration of the education chain of higher vocational colleges and universities and the innovation chain, the professional chain and the industrial chain [9–12]. 12]. The construction of industrial colleges is an innovative move to deepen university-enterprise cooperation, promote industry-teaching collaborative education and solve the problem of mismatch between supply and demand in higher education, and as a new form of promoting the integration of the education chain of higher vocational colleges and universities with the innovation chain and the professional chain with the industrial chain, it greatly promotes the cultivation of high-quality skilled and innovative talents and the reform of the talent cultivation mode of higher vocational colleges and universities [13–16].

The research constructs the evaluation index system for the construction of industrial colleges from the four dimensions of management system, talent cultivation, scientific research innovation and production service. Combining the hierarchical analysis method with fuzzy theory, the improved fuzzy hierarchical analysis method is proposed, and the CRITIC method is used to assign weights to the selected indicators. The evaluation results of the actual case of “Colorful Craftsmen” Auto Repair Industrial College are then measured by using the gray correlation analysis method. Through the selection of relevant variables and the collection of questionnaire data, the influencing factors of talent cultivation development are explored, and through the construction and analysis of multivariate linear regression model, the linear relationship between each variable and the development of talent cultivation is obtained. On this basis, it puts forward the talent cultivation model for the construction of industrial colleges, and builds a multi-dimensional path to jointly promote the development of talent cultivation in industrial colleges by starting from the three links of top-level design of teaching, curriculum teaching and practical teaching.

2

Evaluation of the construction of industrial colleges under the integration of industry and education

Under the perspective of industry-education integration, a reasonable evaluation index system is established by combining the connotation and characteristics of modern industrial colleges, and considering the weight of the index system and the fuzzy and random uncertainty characteristics of the evaluation data, an evaluation model for the construction of industrial colleges is set up and applied to the evaluation of the construction of “Colorful Craftsmen” Automobile Repairing Industrial College.

2.1

Evaluation indicators

Talent cultivation, scientific research innovation and local service are the core functions of institutions of higher education, and the quality of talent cultivation, the level of scientific research innovation and achievement transformation, and the production service capacity under the university-enterprise-industry-university-research collaborative innovation mechanism are the core elements of the performance evaluation of industrial colleges, which reflect the management function, nurturing function, innovation function, and service function of industrial colleges. This paper utilizes the expert consultation method, combines the micro data collected from empirical research, and utilizes SPSS software to conduct independence correlation test on the evaluation indexes, and establishes a set of evaluation index system for industrial colleges with strong applicability. Table 1 shows the evaluation index system for the construction of industrial colleges, including 4 first-level indicators and 16 second-level indicators.

Table 1.

The evaluation index system of the construction of the industry college

Primary index	Secondary index
Management system X1	Construction development X11
	Governance structure X12
	Cooperative enterprise X13
	Professional cluster X14
Talent cultivation X2	Professional construction X15
	Teacher team X16
	Culture pattern X17
	Culture quality X18
Scientific innovation X3	R&D funding X19
	R&D project X20
	R&D results X21
	Resultant transformation X22
Production service X4	Job employment X23
	Cultural communication X24
	Management consultation X25
	Training service X26

2.2.

Evaluation methodology

2.2.1

Improved IF-AHP

In order to clarify the indicator weights, based on the indicators screened in the previous section, an improved strategy combining hierarchical analysis and fuzzy theory is used to propose an innovative application of the IF-AHP fuzzy hierarchical analysis method. The specific solution steps of this improved IF-AHP combined assignment method are described in detail below: 1)

Construction of fuzzy judgment matrix

Suppose Y is an intuitionistic fuzzy set on the argument domain D : Y = {[X,μ₀(x),γ_v(x)]| x ∈ D}, and the rows and columns in the fuzzy judgment matrix M = (m_ij), denoted by i and j, respectively. When denoting the importance of the i th indicator relative to the j th indicator, μ(x) and γ_A (x) are used to denote the degree of affiliation and non-affiliation, respectively, of the x belonging to the intuitionistic fuzzy set Y and satisfying the following conditions: 1 $μ; D \in (0, 1) x \in X \to μ (x) \in (0, 1)$ 2 $γ_{0} : D \in (0, 1) x \in X \to γ_{i j} (x) \in (0, 1)$ 3 $0 \leq μ (x) + γ_{i j} (x) \leq 1$ π_ij(x) is used to scale the degree of hesitancy of x belonging to the intuitionistic fuzzy set Y, for any x ∈ D, there is 0 ≤ π_ij(x) ≤ 1. And is satisfied: 4 $π_{i j} (x) = 1 - μ_{i j} (x) - γ_{i j} (x)$

2)

Consistency test

Constructed fuzzy judgment matrix in order to avoid the emergence of A is more important than B, B is more important than C and C is more important than A need to pass the consistency test, the specific steps are as follows:

(1)

Construct the consistency judgment matrix $M^{'} = (m_{3}^{'})$ , when j > i + 1, take $m_{i j}^{'} = (μ_{i j}, γ_{i j}^{'})$ . When j = i + 1, j = i or i = j + 1, take $m_{g}^{'} = m_{g}$ . when i < j + 1, take $m_{i j}^{'} = (γ_{i j}^{'}, μ_{i j})$ . then: 5 $μ_{i j}^{'} = \frac{\sqrt[j - j - 1]{\prod_{i = i + 1}^{j - 1} μ_{i} μ_{i}}}{\sqrt[j - j - 1]{\prod_{i = i + 1}^{j - 1} μ_{i} μ_{i}} + \sqrt[j - j - 1]{\prod_{i = i + 1}^{j - 1} (1 - μ_{i}) (1 - μ_{i})}}$ 6 $γ_{i j}^{'} = \frac{\sqrt[j - 1]{\prod_{i = 1}^{j - 1} γ_{i j} γ_{i j}}}{\sqrt[j - 1 - 1]{\prod_{i = 1}^{j - 1} γ_{i j} γ_{i j}} + \sqrt[j - 1 - 1]{\prod_{i = 1}^{j - 1} (1 - γ_{i j}) (1 - γ_{i j})}}$

(2)

Apply equation (7) for consistency test: 7 $d (M, M^{'}) = \frac{1}{2 n} \sum_{i = 1}^{\infty} (| μ_{i j} - μ_{i j}^{'} | + | γ_{i j} - γ_{i j}^{'} | + | π_{i j} - π_{i j}^{'} |)$

If d(M′, M) ≤ 0.1, the matrix $\bar{M}$ is considered to pass the consistency test, otherwise the next step will be taken.

(3)

This step does not require re-scoring by the experts, but only modifies the matrix iteration parameters, denoted as σ, σ ∈ [0,1]. calculates the new fuzzy judgment matrix ${\tilde{m}}_{i} = ({\tilde{μ}}_{i}, {\tilde{γ}}_{i})$ ), where: 8 $μ_{i j}^{''} = \frac{{(μ_{i})}^{1 - σ} {(μ_{i j}^{'})}^{σ}}{{(μ_{i})}^{1 - σ} {(μ_{i j}^{'})}^{σ} + {(1 - μ_{i})}^{1 - σ} {(1 - μ_{i j}^{'})}^{σ}}$ 9 $γ_{i j}^{''} = \frac{{(γ_{i j})}^{1 - σ} {(γ_{i j}^{'})}^{σ}}{{(γ_{i j})}^{1 - σ} {(γ_{i j}^{'})}^{σ} + {(1 - γ_{i j})}^{1 - σ} {(1 - γ_{i j}^{'})}^{σ}}$

Substitute the new judgment matrix into Eq. (7) to test $d (\bar{M}, \tilde{M}) < 0.1$ until it passes the consistency test.

3)

Assignment of weights: 10 $H_{i} = \frac{\sum_{j = 1}^{n} μ_{i} - π_{i} / 2}{\sum_{j = 1}^{n} γ_{i j} - π_{i} / 2}$

Calculation of the value of the weight of the j th indicator is calculated according to the formula (11): 11 $ω_{i} = \frac{H_{i}}{\sum_{k = 1}^{n} H_{k}}$

2.2.2

CRITIC method

The CRITIC method, as an objective weighting method, has a significant feature compared with the traditional entropy weighting method and standard deviation weighting method, namely, it incorporates the comparability and potential conflict between evaluation indicators. This feature enables the CRITIC method to assess the objective weights of each indicator in a more comprehensive and detailed way, thus providing more accurate and reliable decision support. Assuming that there are m samples, each with n indicators, the diagnostic indicator data matrix X can be expressed as: 12 $X = [\begin{matrix} x_{11} & x_{12} & \dots & x_{1 n} \\ x_{21} & x_{22} & \dots & x_{2 n} \\ ⋮ & ⋮ & ⋱ & ⋮ \\ x_{n + 1} & x_{n + 2} & \dots & x_{n n} \end{matrix}] n$

The specific calculation steps of the CRITIC method are as follows: 1)

Dimensionless processing:

The dimensionless processing of the indicators is to eliminate the influence due to the difference in the scale, so that the indicators can be compared with each other. In the CRITIC weighting method, forward or reverse processing is usually used to achieve dimensionlessness, while standardization is inconvenient. The reason for this is that if standardization is used, all standard deviations become 1, which means that their volatility is the same, and the degree of importance of each indicator cannot be distinguished, so standardization is not applicable in the CRITIC weighting method.

For positive indicators (e.g., the larger the value of an indicator such as profit, the better): 13 $x_{i j}^{'} = \frac{x_{j} - x_{\min}}{x_{\max} - x_{\min}}$

For negative indicators (e.g., the smaller the value of an indicator such as cost, the better): 14 $x_{i j}^{'} = \frac{x_{\max} - x_{j}}{x_{\max} - x_{\max}}$

The matrix X′ is obtained after normalization from the above equation.

2)

Calculation of indicator variability: 15 ${\begin{array}{l} {\bar{x}}_{j} = \frac{1}{n} \sum_{i = 1}^{*} x_{i j} \\ S_{j} = \sqrt{\frac{\sum_{i = 1}^{*} {(x_{i} - {\bar{x}}_{i})}^{2}}{n - 1}} \end{array}$

3)

Calculation of indicator conflictiveness:

In the CRITIC method, the correlation coefficient is used to measure the correlation between indicators. Correlation coefficient refers to the degree of linear correlation between two indicators, the higher the correlation coefficient, the higher the degree of correlation between the two, that is, the two can reflect similar information to some extent. In the process of assigning weights, the indicator with higher correlation with other indicators may express part of the duplicated information, and the smaller its unique role is, the weight of which should be reduced appropriately. Therefore, in the CRITIC method, the larger the correlation coefficient is, the smaller its weight is in the comprehensive judgment. The specific formula is as follows: 16 $R_{j} = \sum_{i = 1}^{p} (1 - r_{i j})$

4)

Calculation of indicator informativeness: 17 $C_{j} = S_{j} \sum (1 - r_{i j}) = S_{j} \times R_{j}$

The larger C_j is, the greater the role of the j th evaluation indicator in the whole evaluation indicator system, the more weight should be assigned to it.

5)

Calculate the objective weights of the indicators: 18 $ϖ_{j} = \frac{C_{j}}{\sum_{j = 1}^{p} C_{j}}$

2.2.3

Gray correlation analysis method

Gray correlation analysis, as a core application of gray system theory, has the main function of exploring the correlation or similarity between different elements within a system. The method is particularly suitable for dealing with situations containing uncertainty and incomplete information, and realizes the ranking and prioritization of evaluation objects by quantifying the degree of correlation between elements. It provides a scientific basis for decision-making. The basic steps for evaluation using this method are as follows: 1)

Assume that the evaluation index set of the program is Z = (z₁, z₂, ⋯, z₄) and the program set is M = (m₁, m₂, ⋯, m_n), where a_i(i = 1, 2, ⋯ m; j = 1, 2, ⋯ 9) is the value of the evaluation index corresponding to the technical program. Construct the initial decision matrix G: 19 $G = [\begin{matrix} a_{11} & a_{12} & \dots & a_{19} \\ a_{21} & a_{22} & \dots & a_{29} \\ ⋮ & ⋮ & ⋱ & ⋮ \\ a_{m 1} & a_{m 2} & \dots & a_{m i j} \end{matrix}]$

2)

Since multiple computational indicators have different sizes and units, in order to ensure the accuracy and consistency of the calculations, it is necessary to perform a standardized conversion of the initial decision matrix before carrying out the calculations. It can also be called dimensionless processing, and the initial decision matrix G is dimensionless, so as to establish a standardized decision matrix specific calculations as in equations (13) and (14).

3)

In this paper, a combination of fuzzy hierarchical analysis and CRITIC method is used to assign weights to the evaluation indexes, where ω = (ω₁, ω₂, ⋯, ω_ij).

4)

Construct the positive ideal solution $b^{'} = (b_{1}^{'}, b_{2}^{'}, \dots, b_{n}^{'})$ where $b_{i}^{'} = M A X_{i}^{'} b_{i}$ , j = 1, 2, ⋯, 9.

5)

According to the gray system theory, taking the positive ideal solution as the reference sequence and the set of schemes as the compared sequence, then the gray correlation between the i th scheme and the positive ideal solution on the j th attribute is: 20 $ξ_{i} = \frac{\min, \min | b_{i} - b_{i}^{+} | + ρ \max \max | b_{i} - b_{i}^{+} |}{| b_{i} - b_{i}^{+} | + ρ \max, \max | b_{i} - b_{i}^{+} |}, i = 1, \dots, m; j = 1, \dots, n$

Where ρ is the discriminant coefficient, which serves to increase the significance of the difference between the correlation coefficients and is usually taken as ρ = 0.5.

After calculating all the gray correlation coefficients from the above equation, the gray correlation matrix E is derived as follows. 21 $E = [\begin{matrix} ξ_{11} & ξ_{12} & \dots & ξ_{19} \\ ξ_{21} & ξ_{22} & \dots & ξ_{29} \\ ⋮ & ⋮ & ⋱ & ⋮ \\ ξ_{m 1} & ξ_{m 2} & \dots & ξ_{m 9} \end{matrix}]$

6)

The gray correlation coefficients between the i th scheme and the positive and negative ideal solutions in each attribute are linearly combined according to the weights, and the gray correlation r_i between the i th scheme and the positive and negative ideal solutions can be obtained: 22 $r_{i} = \frac{1}{n} \sum_{j = 1}^{n} ω_{j} ξ_{j}, i = 1, \dots, m$

2.3

Evaluation indicator weights

Using the scale method of 1~9 and its reciprocal, experts are organized to rate the importance of each level of evaluation indicators, and the improved IF-AHP method and CRITIC method are used to calculate the weights of evaluation indicators for the construction of industrial colleges. The weights of evaluation indicators of industrial college construction are shown in Figure 1. The weights of Management System X1, Talent Cultivation X2, Scientific Research and Innovation X3 and Production Service X4 are correspondingly 0.245, 0.270, 0.249 and 0.236. Talent Cultivation has the largest weight among the first-level indexes, which indicates that experts generally consider talent cultivation to be the core task in the construction of industrial colleges. Among the second-level indicators, the aspects of job employment X23, achievement transformation X22, construction development X11 and cultivation quality X18 have the highest weights among the first-level indicators, and the weights of the indicators are above 0.08. Such second-level indicator weights fully reflect the experts’ general understanding of the outcome (OBE) orientation of the construction of industrial colleges.

2.4

Empirical analysis

Evaluating the construction of the “Colorful Craftsmen” Automobile Repair Industry College, the evaluation set V was established, and the evaluation of the construction indicators of the industry college was classified into five ratings: excellent, good, medium, low, and very low, corresponding to the grading standards of (4, 5], (3, 4], (2, 3], (1, 2], and (0, 1]. Ten industrial college assessment experts from higher education institutions, government departments, industry associations, and cooperative enterprises were organized to conduct the scoring. According to the experts’ scoring table and the gray correlation analysis method, the comprehensive evaluation results of the construction of “Colorful Craftsmen” Auto Repair Industry College are shown in Figure 2. It can be seen through the comprehensive evaluation that the comprehensive evaluation of the construction of “Seven-Color Craftsmen” Auto Repair Industry College is 3.68 points, which is in the range of average to high level, and the overall situation is developing well.

The management system of “Seven-color Craftsmen” Automobile Repair Industrial College has the highest score of 4.07 for the X1 level indicator, which is excellent, reflecting that the cooperative schools and enterprises of this industrial college have a high degree of coupling and synergy in terms of the common goal of construction, construction and development, governance structure, support, professional clusters, etc., which belongs to the school-enterprise bi-directional dominant modern industrial college with the sustainability of development. It is a two-way school-enterprise-led modern industrial college with sustainable development.

The scores of the first-level indexes of production service X4 and talent cultivation X2 of the industrial college are 3.83 and 3.56 respectively, which are at the level of average to good, and still need to be further improved, among which the enhancement of the faculty X16 and cultivation mode X17 of the industrial college appear to be more important, and their evaluation scores are lower, at 2.46 and 2.73.

The evaluation score of research innovation X3 of the industrial college is 3.29, which is the lowest among all the scores of the first-level indexes, reflecting that the university-enterprise of this industrial college still has big deficiencies in the introduction of scientific research talents, the construction and sharing of scientific research platforms, and the cooperation of scientific research and innovation services.

3

Analysis of factors affecting the development of talent training

The development of talent cultivation in college industrial colleges can promote the deep integration of industry, academia and research. On the basis of assessing the construction level of college industrial colleges, combining the multiple linear regression method to explore in depth the influencing factors of the development of talent cultivation in colleges, academia and research, and to provide references for the subsequent proposal of talent cultivation mode of industrial colleges.

3.1

Scale design

In this study, students’ satisfaction with the education concept, curriculum system, practical teaching quality, enterprise participation, guarantee mechanism, the effect of University-Industry-Research cooperation and the overall development of University-Industry-Research talent cultivation are taken as the reference variables, and a questionnaire with 30 items is designed and distributed to the teachers and students of the “Colorful Craftsmen” College of Automobile Repairing Industry, and 269 questionnaires are recovered as the data set for the analysis of this study. A total of 269 questionnaires were collected as the data set analyzed in this study.

The Cronbach’s coefficient value of the questionnaire data is 0.874, which is greater than 0.8, and the KMO value is 0.865, and it passed the Bartlett’s globular test at the significance level of 0.05. Therefore, it can be inferred that this questionnaire is reliable. Therefore, it can be inferred that the credibility of this questionnaire is good, the evaluation questionnaire has a high degree of internal consistency, and the data analysis of this questionnaire can more accurately and reliably reflect the evaluation status of the students on the effect of university industry-university-research talent cultivation.

3.2

T-test results

The sample t-test results of the variables are shown in Table 2. P<0.01 in the test data means that the mean value of education concept, curriculum system, practical teaching quality, enterprise participation, guarantee mechanism, cooperation effect, and overall satisfaction is 1.497~2.063, which is significantly lower than the test value of 3. It means that the respondents of this research are positive and recognize the evaluation of University-Industry-Research Talent Cultivation Development.

Table 2.

Sample T test results of the variable

Variables	Mean	P	The difference is 95% confidence interval
Variables	Mean	P	Upper limit	Lower limit
Education concept	1.981	0.005	-1.423	-1.236
Curriculum	1.985	0.002	-1.582	-1.103
Practice teaching quality	2.012	0.006	-1.659	-1.019
Enterprise participation	1.691	0.004	-1.524	-1.043
Safeguard mechanism	2.063	0.008	-1.405	-1.101
Cooperative effect	1.862	0.001	-1.468	-1.078
Overall satisfaction	1.497	0.003	-1.514	-1.125
Test value=3

The respondents are more inclined to be satisfied with the evaluation of education concept, curriculum system, practical teaching quality, enterprise participation, guarantee mechanism, cooperation effect and satisfaction with the development of University-Industry-Research Talent Cultivation, and the mean values are significantly lower than the test value of 3 (general), which is a neutral attitude.

3.3

Research methodology

Multiple linear regression (MLR) is a method of statistical analysis that is widely used in various fields. Its purpose is to explore the intrinsic link between an indicator and multiple other indicators and to explain this link by solving for the coefficients of the multiple indicators.

3.3.1

Multiple linear regression models

Multiple linear regression modeling, as the name suggests, examines the relationship between a dependent variable and multiple independent variables, and is similar in principle compared to one-dimensional linear regression, except that it is more computationally complex.

Compared with univariate linear regression, the principle is similar, but the calculation is more complicated. If we assume that the dependent variable is y, and we need to investigate the linear relationship between the dependent variable y and the nn independent variables x₁, x₂, …, x_n, its general multiple linear regression formula is shown in equation (23): 23 $y = \sum_{i = 1}^{n} b_{i} x_{i} + b_{0} + ε$

Where: b_i denotes the partial regression coefficient of its corresponding independent variable in the sample, which means the amount of change in the dependent variable y averaged over a unit change in the independent variable. b₀ is a constant that represents the intercept. ε is the random error after removing the effect of the n independent variables on the dependent variable y, also known as the residual, which is an unobservable random variable. If there are m sets of observations, then substituting them into the general formula yields the formula shown in Eq. (24): 24 ${\begin{matrix} y_{1} = b_{0} + b_{1} x_{11} + b_{2} x_{12} + \dots + b_{n} x_{1 n} + ε_{1} \\ y_{2} = b_{0} + b_{1} x_{21} + b_{2} x_{22} + \dots + b_{n} x_{2 n} + ε_{2} \\ ⋮ \\ y_{m} = b_{0} + b_{1} x_{m 1} + b_{2} x_{m 2} + \dots + b_{n} x_{m n} + ε_{m} \end{matrix}$

Write it in vector form as shown in equation (25): 25 ${\begin{matrix} y = {[\begin{array}{l} y_{1} & y_{2} & \dots & y_{m} \end{array}]}^{T} \\ X = [\begin{matrix} 1 & x_{11} & x_{12} & \dots & x_{1 n} \\ 1 & x_{21} & x_{22} & \dots & x_{2 n} \\ ⋮ \\ 1 & x_{m 1} & x_{m 2} & \dots & x_{m n} \end{matrix}] \\ b = {[\begin{array}{l} b_{0} & b_{1} & \dots & b_{n} \end{array}]}^{T} \\ ε = {[\begin{array}{l} ε_{1} & ε_{2} & \dots & ε_{m} \end{array}]}^{T} \end{matrix}$

Then equation (24) can be written as equation (26): 26 $[\begin{matrix} y_{1} \\ y_{2} \\ ⋮ \\ y_{m} \end{matrix}] = [\begin{matrix} 1 & x_{11} & x_{12} & \dots & x_{1 n} \\ 1 & x_{21} & x_{22} & \dots & x_{2 n} \\ ⋮ \\ 1 & x_{m 1} & x_{m 2} & \dots & x_{m n} \end{matrix}] [\begin{matrix} b_{0} \\ b_{1} \\ ⋮ \\ b_{n} \end{matrix}] + [\begin{matrix} ε_{1} \\ ε_{2} \\ ⋮ \\ ε_{m} \end{matrix}]$

Eq. (26) can then be expressed as Eq. (27): 27 $y = X b + ε$ where the vector of partial regression coefficients b is the unknown parameter and s is the residuals. In general X is referred to as the information matrix of order m *(n + 1) and X is assumed to be column full rank, i.e: 28 $r a n k (X) = 1 + n$

Eq. (28) represents the rank of matrix X and Eq. (27) is a classical linear algebra formula. The objective is to solve for the vector of partial regression coefficients b in the model.

3.3.2

Least Squares

The purpose of multiple linear regression is to solve an optimal equation, through which the value of the dependent variable y can be predicted with known information about the independent variable X. In this optimal equation, “optimality” is reflected in the regression coefficients. Generally, the regression coefficients of the equation can be solved by using the sum of squares of the minimization errors. If the sum of differences is used as the regression coefficient solution method, in the solution process, due to the existence of positive and negative values of the difference, there will be positive and negative cancel each other out, which will have a negative impact on the regression results, so the sum of squares of the error is generally used. The sum of squared errors is shown in equation (29) below: 29 $Q (b) = \sum_{i = 1}^{m} {((b_{0} + b_{1} x_{i 1} + b_{2} x_{i 2} + \dots + b_{n} x_{i n}) - y_{i})}^{2}$ where Q(b) is the sum of squares of the errors between the predicted and actual values. Then multivariate linear regression is to solve for the value of the parameter vector b such that the value of Q(b) in equation (29) above is minimized. The process of minimizing the Q(b) function as a constraint to find the optimal solution to the parameter vector is the least squares algorithm.

Defining $y_{i}^{'}$ as the predicted value of the model corresponding to the actual result $y_{i}^{'}$ with the following expression (30), equation (29) can be written as the following equation (31): 30 ${\begin{array}{l} y_{i}^{'} & = b_{0} + b_{1} x_{i 1} + b_{2} x_{i 2} + \dots + b_{n} x_{i n} \\ y^{'} & = {[\begin{array}{l} y_{1}^{'} & y_{2}^{'} & \dots & y_{m}^{'} \end{array}]}^{T} = X b \end{array}$ 31 $h (b) = \sum_{i = 1}^{m} {(y_{i}^{'} - y_{i})}^{2}$ where h(b) is defined as being a function so that solving for the Q(b) minimum can be transformed into solving for the h(b) minimum. Combining Eq. (25), Eq. (30) and Eq. (31) yields Eq. (32): 32 $h (b) = \sum_{i = 1}^{m} (y_{i}^{'} - y_{i})^{2} = (y - y^{'})^{T} (y - y^{'}) = (y - X b)^{T} (y - X b)$

When y′ and y are infinitely close, then h(b) is equal to a very small value of δ, the following formula (33), at this time, its two sides of the parameter b respectively to obtain the derivation of the formula (34), the expansion of which can be obtained to obtain the final regular equation as in Eq. 33 $h (b) = (y - X b)^{T} (y - X b) = δ$ 34 $2 X^{T} (y - X b) = 0$ 35 $b = {(X^{T} X)}^{- 1} X^{T} y$

Where X^T is the transpose matrix of matrix X and(X^TX)⁻¹ is the inverse matrix of X^TX, it is also known that this method must be applied in the presence of the inverse matrix of X^TX. After obtaining the formal equation shown in Eq. (35), the optimal parameters can be solved by bringing in the known parameters, and finally the model of the multi-distance linear regression equation can be determined and the prediction can be made by this model.

3.3.3

Performance testing of the model

A series of tests are needed after the model is built, such as the performance of the model. The tests of regression model include: goodness-of-fit test, significance test of regression equation and significance test of regression coefficient. Among them, the goodness-of-fit test, also known as the coefficient of determination, is a frequently used test of regression model performance in research. The goodness of fit, i.e., the closeness of the predicted value of the model to the true value of the sample, is expressed by the coefficient of determination r², which is solved by the formula in equation (36): 36 ${\begin{matrix} S S R = \sum_{i = 1}^{n} {(y_{i} - \bar{y})}^{2} \\ S S E = \sum_{i = 1}^{n} {(y_{i} - y_{i})}^{2} \\ S S T = \sum_{i = 1}^{n} {(y_{i} - \bar{y})}^{2} \\ r^{2} = \frac{S S R}{S S T} = \frac{\sum_{i = 1}^{n} {(y_{i} - \bar{y})}^{2}}{\sum_{i = 1}^{n} {(y_{i} - \bar{y})}^{2}} \end{matrix}$

Where y_i. is the true value of the first i samples, ${\hat{y}}_{i}$ is the predicted value of the first i samples, and $\bar{y}$ is the mean of the true values of the n samples.SSR is known as the sum of squares of regression, which expresses the deviation of the fitted predicted values from the mean, SSE is known as the sum of squares of residuals, and SST = SSR + SSE is known as the sum of squares of the total deviations. From equation (36), r² takes values in the range [0, 1] and depends on the value of SSR. The closer the value of r² is to 1, i.e., the larger the value of SSR, the better the goodness of fit of the multiple linear regression equation is indicated, and vice versa.

3.4

Regression model analysis

3.4.1

Correlation analysis

Pearson correlation analysis was performed on the data collected from the scale. Mainly examining the correlation between the variables, the reference variables were analyzed by PP charts, which showed normal distribution, so the correlation analysis chose the Pearson statistical analysis method, and the specific data of the Pearson correlation analysis are detailed in Table 3, *** indicates p < 0.01, ** indicates p < 0.05, and * indicates p < 0.1.

Table 3.

The specific data of the Pearson correlation analysis

	1	2	3	4	5	6	7
1.Education concept	1
2.Curriculum	0.759***	1
3.Practice teaching quality	0.764***	0.817***	1
4.Enterprise participation	0.609**	0.685	0.759***	1
5.Safeguard mechanism	0.748*	0.710*	0.712**	0.775*	1
6.Cooperative effect	0.774**	0 797***	0.838***	0.666***	0.785**	1
7.Overall satisfaction	0.686***	0.816***	0.812***	0.695***	0.781***	0.692***	1

There is a significant correlation (p < 0.01) between the educational philosophy, curriculum system, practical teaching quality, enterprise participation, guarantee mechanism, cooperation effect and satisfaction with the development of industry-university-research talent cultivation, reserving participation in the subsequent multi-factor analysis.

3.4.2

Multi-factor analysis

Taking education concept, curriculum system, practice teaching quality, enterprise participation, safeguard mechanism, and cooperation effect as independent variables, and satisfaction with the development of industry-university-research talent cultivation as dependent variables, multiple linear regression analysis was conducted, and the results of the multifactor analysis are shown in Table 4. Satisfaction with the development of industry-university-research talent cultivation=0.178* Education concept+0.228* Curriculum+0.344* Practice teaching quality+0.352* Enterprise participation+0.202* Safeguard mechanism+0.365* Cooperative effect, it can be seen from the equation: industry-university-research cooperation effect> Enterprise participation> Practice teaching quality> Curriculum system> Safeguard mechanism> Education concept. Based on the above analysis, it can be concluded that all six variables have an important impact on the development of industry-university-research talent cultivation in local colleges and universities in different degrees.

Table 4.

The result of multi-factor analysis

Variables	Unnormalized coefficient	S.E.	Normalized coefficient	t	Sig.	VIF
(Constants)	-0.214	0.672		-2.795	0.685	3.619
Education concept	0.178	0.021	0.155	1.738	0.008	2.678
Curriculum	0.228	0.052	0.185	3.709	0.004	4.728
Practice teaching quality	0.344	0.094	0.363	2.732	0.007	6.844
Enterprise participation	0.352	0.075	0.374	4.776	0.005	5.672
Safeguard mechanism	0.202	0.086	0.283	3.771	0.003	4.812
Cooperative effect	0.365	0.081	0.383	2.741	0.004	3.625
F			374.583
P			<0.01
R²			0.758
Factor: the satisfaction of talent cultivating development

4

Talent training model under the integration of industry and education

As a product of multi-party collaborative education mode in the context of industry-teaching integration, the Industrial College has innovated the depth, breadth and strength of university-enterprise cooperation, which makes high-quality enterprise resources participate in the cultivation of talents in colleges and universities, and deeply promotes the collaborative education in industry-academia cooperation and at the same time pushes forward the reform of talent cultivation in colleges and universities by the latest needs of industrial and technological development. Through the analysis of the influencing factors of talent cultivation, the talent cultivation of industrial colleges emphasizes the effect of enterprise cooperation, the quality of practical teaching, the construction of the curriculum system and teaching philosophy, based on which the talent cultivation mode of industrial colleges is discussed. The talent cultivation model of industrial colleges is shown in Fig. 3, which is designed from the three links of teaching top level, curriculum teaching and practical teaching.

4.1

Teaching and Learning Top-Level Design Sessions

4.1.1

Demand for Development of the Research Industry

Before the development and revision of the training program, professional experts and teachers of professional courses are organized to discuss the development trend of the relevant industries, assess the development of the industry’s demand for talents, develop and revise the professional training program accordingly, and enhance the degree of fit between the training of talents and the industrial demand.

4.1.2

Joint development of training programs

The school encourages professional and enterprise related units to participate in the development and revision of the training program, by inviting enterprise units to participate in the training program discussion, the industrial development of professional knowledge of talents, the ability of the demand for talents directly into the training program, and targeted setting of relevant courses, school-enterprise co-construction of courses.

4.1.3

School-enterprise co-construction of special courses

The joint construction of industry-specific courses by universities and enterprises is an important part of the “integration of industry and education” in educating people. Colleges and universities build special courses with relevant professional companies according to the needs of industrial development, and invite industry experts to give lectures, so that students can face the industry, understand what the industry needs, and learn in a more targeted way, so as to cultivate professionals required by the industry development. Cooperation between colleges and industries helps optimize the allocation of educational resources. By sharing industrial resources, such as laboratories, equipment, technological resources and funds, colleges and universities are able to improve the quality of education and provide students with better learning and practice conditions.

4.2

Classroom teaching sessions

4.2.1

Problem-oriented introduction

Relying on scientific research projects, lecturers bring the problems and challenges faced in the professional field into the classroom through discussions and assignments to inspire students’ thinking, stimulate their enthusiasm for solving the frontline problems of industrial development, and better learn the basic theoretical knowledge. Meanwhile, to strengthen practical education, the university can provide students with practical projects and internship opportunities through internships with cooperative enterprises, participation in industrial research projects, participation in social welfare projects, etc., so as to cultivate students’ practical operation and problem solving ability and enhance their adaptability in the workplace.

4.2.2

Inviting industry personnel to give lectures

For the content of specific links, industrial personnel are invited into the classroom to explain the status quo of industrial development, industrial development needs, industrial development dilemmas, etc., so that students can participate in the discussion in the classroom. It not only enriches the teaching method of the course, but also increases the interest of the classroom. At the same time, it also enhances students’ understanding of industrial development and stimulates their enthusiasm for solving practical problems.

4.3

Practical Teaching Sessions

4.3.1

Building an experimental platform

In order to overcome the disconnection between students’ theoretical learning and practical ability, and also for better understanding and mastering of theoretical knowledge, oriented by the industrial application demand, combined with the professional actuality, self-built or co-created with enterprises to create a simulation virtual experimental platform, to enhance the students’ ability to solve practical complex problems. Pioneering simulation experimental platform, for students in school to realize the combination of theoretical knowledge learning and solving the needs of practical application of engineering to provide a nurturing platform support.

4.3.2

Conducting practical courses

In the practical course sessions, students are arranged to go into enterprises according to the actual situation of their specialties to understand the development of the industry. Arrange students to go into the front line of engineering to understand the progress of engineering technology and the challenges they face, and discuss the possible ways to solve them. Arrange for students to go into the community to understand the current situation of the community and the difficulties of governance, and discuss coping strategies in the context of their specialties, so as to continuously expand the boundaries and effectiveness of the “integration of industry and education”.

5

Conclusion

Based on the perspective of industry-education integration, the study constructs an evaluation index system for the construction of industrial colleges, and evaluates the construction of industrial colleges on the example of “Colorful Craftsmen” Auto Repair Industrial College. Then we select several variables to explore the influencing factors of talent cultivation development in industrial colleges, and then put forward the talent cultivation model. The main research results are as follows: 1)

Talent cultivation has the greatest weight (0.270) in the construction of industrial colleges, which is the key index to measure the construction level of industrial colleges. Specifically, the weights of job employment, achievement transformation, construction development and cultivation quality are all greater than 0.08, which are important factors in the construction and development of industrial colleges. The construction of “Colorful Craftsmen” Auto Repair Industrial College is at a good level, with an evaluation score of 3.68, of which the evaluation score of management system is the highest (4.07), reaching an excellent level, and the evaluation scores of production service, talent cultivation and scientific research and innovation are at a good level, with evaluation scores of 3.83, 3.56 and 3.29. There is still a lot of room for upward mobility in the level of research and innovation in this college of the automotive industry.

2)

All the variables selected in this paper have significant positive correlation (p < 0.01) to the talent cultivation development of industrial colleges. The regression coefficients of the effect of industry-university-research cooperation, enterprise participation and practical teaching quality are 0.365, 0.352 and 0.344 correspondingly, and the effect on the development of talent cultivation in industrial colleges is more obvious.

3)

Through the analysis of influencing factors, the talent cultivation mode of industrial colleges is proposed from the teaching top-level design link, course teaching link and practice teaching link, so as to better cultivate high-quality talents adapted to the needs of the society and to promote economic development and social progress.

Sprache:: Englisch

Zeitrahmen der Veröffentlichung:: 1 Hefte pro Jahr
Fachgebiete der Zeitschrift:: Biologie, Biologie, andere, Mathematik, Angewandte Mathematik, Mathematik, Allgemeines, Physik, Physik, andere

Zeitschrift RSS Feed

Research on the Construction of Industrial College and Talent Cultivation Mode under the Perspective of Industry-Education Integration--Taking “Seven-Color Craftsmen” Auto Repair Industrial College as an Example

Yuexiang Chen

Zhaoxin Wang

Ni Xu

Online veröffentlicht: 24. Sept. 2025

Eingereicht: 11. Jan. 2025

Akzeptiert: 19. Apr. 2025

DOI: https://doi.org/10.2478/amns-2025-0948

SchlüsselwörterIF-AHP, CRITIC method; Gray correlation, Multiple linear regression, Evaluation index system, Industry-teaching integration

© 2025 Yuexiang Chen et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Schlüsselwörter
IF-AHP, CRITIC method; Gray correlation, Multiple linear regression, Evaluation index system, Industry-teaching integration