Research on the Innovative Path of Mental Health Education for College Students in Higher Vocational Colleges and Universities Aided by Computer and Intelligent Technology

With the development of human society, the rapid changes in science and technology, the accelerating process of economic globalization and the intensification of social competition, people today are facing unprecedented impacts and challenges. Such impacts and challenges, on the one hand, create better conditions for the promotion of comprehensive human development and, on the other hand, make people face more and more psychological pressure and complex psychological problems. People have gradually begun to recognize the important role of psychological and social factors in health and disease and their mutual transformation that cannot be ignored [1-2]. Today’s world is in an era of fierce competition. The so-called competition, in essence, is the competition of talent, and college students are the future builders of the motherland, so the field of higher education shoulders the important task of cultivating high-quality talents for the 21st century [3-4]. Practice shows that in order to effectively fulfill this important task, it is necessary to improve the overall quality of students from a systematic point of view.

And mental health education is an important foundation and prerequisite for comprehensively improving the overall quality of students. Modern education theory holds that education is not only to impart knowledge and skills but also to equip students with good psychology and shape a healthy, harmonious and stable personality for future comprehensive development [5-7]. Improving the mental health of college students is an important guarantee for college students to move towards modernization, the world and the future, and to build the motherland and the homeland [8]. Mental health education plays a very important role in cultivating college students’ healthy psychology, developing their psychological potential, enhancing their adaptability and perfecting their personality. With the acceleration of the social rhythm and the intensification of competition, the market economy has put forward higher requirements for talent, and being able to adapt to this fiercely competitive environment, being able to face all kinds of unsatisfactory predicaments, always maintaining an optimistic attitude, and correctly looking at the problems arising in the society is a higher requirement for the quality of mental health of college students [9-10]. Strengthening the research of mental health education for college students, guiding them to maintain physical and mental health, preventing various mental illnesses, and cultivating high-quality qualified talents for society is the call of the progress of the times. Mental health education has become a difficult, complex and significant task that affects the lifelong development of individual college students and the cultivation of modernization and construction talents by the state [11].

Although students entering university tend to be rational and mature in terms of knowledge and mind, due to the lack of deep understanding of reality and experience, there is a certain vulnerability in the psychological aspect, which leads to the fact that they are prone to mental health problems due to some bad experiences [12-13]. Under the current situation where the pace of life and the pressure of study continue to increase, mental health problems have obviously become a prominent problem in the college student population. Therefore, more and more colleges and universities have incorporated mental health education into the education system of college students, hoping to enhance the ability of college students to cope with and crack psychological problems through professional and targeted mental health education. However, the diversity of students’ psychological status and the lack of resources for mental health education teachers make it difficult to effectively meet students’ individualized mental health education needs and mental health education is often reduced to a kind of public knowledge course [14-15]. In such a situation, it is necessary to use technology to improve the ability of college teachers to screen, locate, and analyze the mental health problems of college students so that teachers can focus more practice and energy on helping students to crack psychological problems. Artificial intelligence and big data, as emerging technological content, can precisely meet the needs in this regard [16].

In today’s society, the healthy growth of college students is not only limited to academic achievement but also includes the overall development of body and mind and good social adaptation. Among them, psychological health is the core element of the overall healthy growth of college students. Literature [17] used an electronic questionnaire to examine the mental health status of students at KU Leuven University in Belgium and found that one-third of them had mental health problems, and pointed out that these mental health problems included both internalized and externalized types and both of them had an impact on academic functioning. Literature [18] combined binary logistic regression analysis methods, demographics and other methods in order to investigate the current status of the use of mental health services on campus as well as trends. The study found that the trend of the use of mental health services among students has gradually increased and pointed out that strengthening the examination of healthcare-related issues can promote the further rational use of mental healthcare resources by college students. Literature [19] on the race and mental health needs of the relevant studies learned that the prevalence of psychological problems of different races is small, but the use of mental health services varies greatly. People of color, Asian students, and other psychological service needs significantly lower than white people. Literature [20], based on a longitudinal survey and structured questionnaire method, examined how the epidemic affects the mental health of college students. Based on the results of the study was informed that the severity of the epidemic affects the quality of sleep and the mood of the students and that appropriate daily sports and adequate sleep can effectively alleviate this mental health problem. Literature [21] based on information and data from the Northern Ireland College Student Mental Health Research Project shows that one in ten newly enrolled students have received mental health treatment, and one in five said they did not want mental health services. The study provides an important reference for educational institutions to address issues related to students’ mental health.

Literature [22] uncovered and analyzed clinical data on the mental health of students in higher education, noting that social anxiety and depression have become the most prevalent mental health problems among students and that the number of students seeking support from mental health services is on the rise. Literature [23] in the context of the epidemic to carry out research related to the mental health problems of college students pointed out that the mental health of students who already have mental health problems improved, while the mental health of the students in the isolated environment of the epidemic, some psychological problems, that colleges and universities need to provide support for students with mental health problems, as well as to provide psychological prevention and intervention for normal students, in order to reduce students’ likelihood of psychological problems arising and worsening. Literature [24] elucidated that economic development and educational expansion, the affluence and education of the country’s people have increased to a certain extent, and the pressure on students’ lives and learning has also skyrocketed, which has also triggered more mental health problems. The current studies on students’ mental health point to a rising trend in college students’ mental health problems, and there are some analyses and even contradictions in the studies on the mental health problems brought about by the epidemic, arguing that future studies on students’ mental health should be more in-depth and scientific and that the direction should point to the underlying logic of the increase in students’ mental problems.

Artificial intelligence and computer-aided technology, as emerging technologies that have attracted much attention in recent years, have a broad application and development space in improving colleges and universities to carry out intelligent and refined education. The application of artificial intelligence and computer-aided technology to the mental health education practice of college students has a very positive effect on the accurate discovery and scientific solution of college student’s mental health problems. Literature [25] systematically reviewed the research literature related to mental health education and found that role-playing strategies used in mental health education can help to improve the effect of mental health and, at the same time, improve the communication ability and empathy of students. Literature [26] reviewed the treatment strategies for mental health problems, pointed out that existing mental health treatments generally reduce the incidence of mental problems and reduce the adverse consequences caused by mental health problems, and concluded that there is a need to further bridge the gap between theory and mental health clinical practice to improve the effectiveness of mental health interventions. Literature [27] describes a whole-school mental health intervention approach with four interrelated cyclical processes to fit this mental health intervention approach, which somewhat optimizes and improves this mental health intervention. Literature [28] examined the impact of clinical simulation strategies on mental health practitioners and combined them with evidence from practice feedback. It was shown that clinical simulation strategies were effective in improving students’ mental health intervention skills and refining students’ mental health knowledge. Literature [29] comprehensively analyzes the process of change in the practice of student mental health interventions and educational integration and discusses it in detail with practical cases, pointing out that the school mental health support system effectively reduces the incidence of mental illnesses in students as well as students’ mental health problems due to academic failures and proposes to strengthen the monitoring of students’ mental health as well as to further coordinate the joint school-community mental health interventions. In the research literature on mental health education, many effective interventions for mental health education in colleges and universities have been put forward. It is believed that these methods of mental health education can be analyzed and optimized in depth in order to improve the effectiveness of mental health interventions. At the same time, the underlying logic of the functioning of mental health interventions can be analyzed comprehensively in order to deepen the understanding of the effectiveness of mental health interventions.

The article firstly proposes a perfect method to judge the mental health status of college students in higher vocational colleges, establishes a multilevel fuzzy comprehensive judgement model, and briefly describes the principle of the established judgement model. Then, a mental health prediction model for college students in higher vocational colleges based on the BP algorithm is designed. Specifically, the optimized BP algorithm is selected to establish the mental health prediction model, and its system structure and function are designed in detail. By establishing the factor set of the fuzzy comprehensive judgement model of college student’s mental health status in higher vocational colleges and using the set-value statistics iterative method to calculate the degree of affiliation of each factor, the obtained degree of affiliation is more in line with the current situation of college students’ mental health in higher vocational colleges and universities. Then, the performance analysis of the design network prediction model designed in this paper is carried out to verify the prediction effect of the prediction model of college student’s mental health status in higher vocational colleges designed in this paper by comparing the practical application of the constructed prediction model with the existing measurement methods. Finally, based on the experimental results, this paper proposes several countermeasures for the psychological health intervention of college students in higher vocational colleges.

2

Method

2.1

Mental health status judgement model

In conducting the mental health testing system, this paper adopts the fuzzy comprehensive judgement method, in view of the complexity and ambiguity of psychological problems, and introduces the multi-level comprehensive judgement model in the comprehensive judgement to study the mental health status of these college students [30].

2.1.1

Establishment of factor sets and rubric sets

The creation of factor set $U = {u_{1}, u_{2}, u_{3}, u_{4}, u_{5}, u_{6}, u_{7}, u_{8}, u_{9}}$ , i.e. a total of 9 factors associated with the judged, refers to the 9 factors referred to in the SCL-90 scale, i.e. somatisation manifestations, obsessive-compulsive symptom manifestations, interpersonal sensitivity manifestations, depressive manifestations, anxiety manifestations, hostile manifestations, fearful manifestations, paranoia manifestations, and psychoticism manifestations, which are indicated by u_i.

Establish the set of rubrics as $V = {v_{1}, v_{2}, v_{3}, v_{4}}$ , i.e., let there be 4 possible rubrics, where good mental health is represented by v₁, normal mental health is represented by v₂, mildly abnormal mental health is represented by v₃, and severely abnormal mental health is represented by v₄, which can be directly translated as V = {good mental health, normal mental health, mildly abnormal mental health, and severely abnormal mental health}.

2.1.2

Establishment of a multilevel fuzzy comprehensive judgement model

Since the decision making of mental health status is more complex and involves more factors, the algorithm has problems in determining, first of all, how the weights of the factors are determined, and what kind of method will be used to be objective and real. Furthermore is that if the weights of the factors are determined, we are also controlled by the constraints of normalization so that the sum of the weights of the factors is 1, so that the weights assigned to each of the 9 factors will be small. Then through the “ ∧” and “∨” operations, the vector B will be synthesized, and the process of multiple operations leads to b_j are very small: $b_{j} \leq V_{i = 1}^{n} a_{i}$ such an important factor weights will be “divided”, so that much information The real results are not evaluated.

Therefore, we have also proposed the multilevel Fuzzy model. Its basic idea is that, when determining a large concept, its determinants will often be many so that these factors are processed according to the same attribute, similar or similar to be classified as a class, weights are assigned, and then a comprehensive judgement is made, and the same determinants in each class depend on many sub-factors or low-level factors, then the low-level sub-factors are also comprehensively assessed. This is divided level by level [31]. For example, the evaluation of school spirit, research, etc., in school evaluation can be subdivided into sub-factors such as teaching, faculty, etc. From this, we propose a multi-factor Fuzzy integrated decision-making model, i.e., a multi-level integrated judgement model. 1)

Implementation steps of the multilevel model (1)

The factor set $U = {u_{1}, u_{2}, \dots, u_{n}}$ is subdivided and grouped into a subset or a unit according to like characteristics, which is divided into s subsets $U_{i} = {u_{i 1}, u_{i 2}, \dots, u_{i n_{i}}}$ , of which i = 1, 2, …, s satisfies the condition: (a)

$\sum_{i = 1}^{S} n_{i} = n$

(b)

$U_{i = 1}^{s} U_{i} = U$

(c)

U_i ∩ U_j = ∅, i ≠ j

(2)

When making a judgement, each sub-factor set is seen as independent and can be combined separately for decision making.

Let $V = {v_{1}, \dots, v_{m}}$ be a subset of the set of decision outcomes and $A_{i} = (a_{i 1}, a_{i 2}, \dots, a_{i n_{i}})$ be the weights assigned to each factor in U_i.

All the weights satisfy the sum 1 law, i.e. $\sum_{i = 1}^{n_{i}} a_{i 1} = 1$ .

Get R_i If it is a one-factor matrix, then the resulting judgement vector is the judgement vector of level 1. $B_{i} = A_{i} \circ R_{i} = (b_{i 1}, b_{i 2}, \dots, b_{i m})$ , where i = 1, 2, …, s.

(3)

Continue to do the division by treating each U_i as a factor and noting: (1) $U ≜ {U_{1}, U_{2}, \dots, U_{s}}$

Construct u becomes the set of factors. The following equation can represent the one-factor decision matrix: (2) $R = [\begin{matrix} B_{1} \\ B_{2} \\ ⋮ \\ B_{i} \end{matrix}] = [\begin{matrix} b_{11} & b_{12} & \dots & b_{1 m} \\ b_{21} & b_{22} & \dots & b_{2 m} \\ ⋮ & ⋮ & ⋮ \\ b_{s 1} & b_{s 2} & \dots & b_{s m} \end{matrix}]$

The formula shows that U consists of a number of U_i, with each U_i reflecting the status of one of the attributes of U. Weights are then assigned to each U_i according to its importance: (3) $A = (a_{1}, a_{2}, \dots, a_{s})$

This makes it easy to derive a secondary judgement vector model. (4) $B = A \circ R = (b_{1}, \dots, b_{m})$

The second level judgement model is shown in Figure 1, which gives an intuitive explanation of the second level decision model.

If each sub-factor set U_i is still large, containing multiple types and multiple numbers of factors, then we divide U_i further, and the result will be a tertiary model, a quaternary model, a quintenary model, and so on. 2)

Determination of weights

A person’s mental health status is often to be defined by psychologists or psychiatrists, coupled with the role of different factors on different patients are not the same, so often the actual situation is determined by the expert survey estimation method to determine the weights or by the experts of the experience of the estimation or determination, but the impact of this paper’s factors more than 3, it is very difficult for any expert to determine the degree of influence of each factor that is the degree of affiliation. So in the determination of the weights, we use the set-value statistical iterative method introduced earlier for the allocation of weights. (1)

Select the p₁ = q factors in U that he considers most important to get a subset of U: (5) $U_{1}^{(j)} = {u_{i 1}^{(j)}, u_{i 2}^{(j)}, \dots, u_{i q}^{(j)}} \subset U$

(2)

Choose 2q factors from U, so that p₂ = 2q, these 2q factors are the most important, and then construct a subset of U, $U_{2}^{(j)} = {u_{i 1}^{(j)}, \dots u_{i q}^{(j)}, \dots u_{i q + 1}^{(j)}, \dots, u_{i q}^{(j)}} \supset U$ and so on.

(3)

Pick sq factors in U such that p_s = sq, which sq is the factor he considers most important, and again construct a subset of U: (6) $U_{s}^{(j)} = {u_{i 1}^{(j)}, u_{i 2}^{(j)}, \dots, u_{i s q}^{(j)}} \supset U_{s - 1}^{(i)}$

Construct until the natural number i satisfies n = iq + r, 1 ° r ° q, then terminate the iterative process, which is step i + q: (7) $(i + 1) F e t c h U_{i + 1}^{(j)} = U$

Then, calculate the coverage frequency of u_i, i = 1, …, n: (8) $m (u_{i}) = \frac{1}{k (i + 1)} \sum_{s = 1}^{i + 1} \sum_{j = 1}^{h} C U_{s}^{(j)} (u_{i})$

(C is the eigenfunction), gauge normalised to give $a_{i} = m (u_{i}) / \sum_{i = 1}^{n} m (u_{i})$ , and finally the weight assignment: (9) $A = (a_{1}, a_{2}, \dots, a_{n})$ 3)

Establishing a second-level fuzzy comprehensive judgement model for students’ mental health status

We divide the above factor set U into levels, which is divided into two levels, and the second level is divided into three factor sets according to the characteristics of the factors, and the fuzzy comprehensive judgement is carried out within each level. The set of rubrics remains unchanged. The 9 factors are divided into three factor sets as follows: (1)

First level factor set: (10) $U = {U_{1}, U_{2}, U_{3}}$

where:

U₁ represents the student’s own physical signs.

U₂ represents external interactions with people.

U₃ represents inner personality and emotions. (2)

Second level factor set: (11) $\begin{array}{l} U_{1} = {u_{1}, u_{5}, u_{9}} \\ U_{2} = {u_{3}, u_{6}, u_{7}} \\ U_{3} = {u_{2}, u_{4}, u_{8}} \end{array}$

The factor sets in U₁ are: u₁ Somatisation factor, u₅ Anxiety factor, u₉ Psychosis.

The factor set in U₂ is: u₃ Interpersonal sensitivity, u₆ Hostility factor, u₇ Fear factor.

The factor set in U₃ is: u₂ Obsessive-compulsive factor, u₄ Depression factor, u₈ Paranoia factor.

Secondary Evaluation Matrix: (12) $U_{1} = [\begin{array}{l} u_{1} \\ u_{5} \\ u_{9} \end{array}] U_{2} = [\begin{array}{l} u_{3} \\ u_{6} \\ u_{7} \end{array}] U_{3} = [\begin{array}{l} u_{2} \\ u_{4} \\ u_{8} \end{array}]$ (3)

For each subset of factors that has been constructed, a combined decision judgement is made, $V = {v_{1}, \dots, v_{4}}$ is the decision set, $A_{i} = (a_{i 1}, a_{i 2}, \dots, a_{i m})$ is the assigned weights of the factors in U_i, and the sum of the weights satisfies $\sum_{i = 1}^{n_{i}} a_{i 1} = 1$ . If R_i is a one-factor matrix, the result of the construction is a one-level judgement vector. $B_{i} = A_{i} \circ R_{i} = (b_{i 1}, b_{i 2}, \dots, b_{i m})$ , where i = 1, 2, 3 (13) $\begin{array}{rcl} B_{1} = A_{1} \circ R_{1} \\ B_{2} = A_{2} \circ R_{2} \\ B_{3} = A_{3} \circ R_{3} \end{array}$

(4)

The second level of the judgement matrix, following the idea of the previous step, considers each U_i as a factor denoted as $u ≜ {U_{1}, U_{2}, U_{3}}$ , and so a further factor set u is constructed. u The one-factor decision matrix is: (14) $R = [\begin{array}{l} B_{1} \\ B_{2} \\ B_{3} \end{array}] = [\begin{array}{l} b_{11} & b_{12} & b_{13} & b_{14} \\ b_{21} & b_{22} & b_{23} & b_{24} \\ b_{31} & b_{32} & b_{33} & b_{34} \end{array}]$

Where U consists of a number of U_i, and each U_i embodies some attribute of U. Weights can be assigned by the set-value statistical iteration method: (15) $A = (a_{1}, a_{2}, a_{3})$

Using the fuzzy matrix synthesis operation, the judgement vector for the second level was then obtained: (16) $B = A \circ R = (a_{1}, a_{2}, a_{3}) \circ [\begin{array}{l} b_{11} & b_{12} & b_{13} & b_{14} \\ b_{21} & b_{22} & b_{23} & b_{24} \\ b_{31} & b_{32} & b_{33} & b_{34} \end{array}] = (b_{1}, b_{2}, b_{3}, b_{4})$

Finally, according to the principle of maximum affiliation, corresponding to the result set $(b_{1}, b_{2}, b_{3}, b_{4})$ of the fuzzy comprehensive judgement, the results of the mental health status of university students were obtained, i.e. (Good, Normal, Mildly Abnormal, Severely Abnormal).

2.2

Neural network prediction model

This section proposes a prediction model based on the BP algorithm, which is used to predict and analyse the mental health status of college students in higher education institutions.

2.2.1

BP algorithm

The BP algorithm is a class of guided learning algorithms for the learning of weights and thresholds in BP nets, which systematically solves the problem of learning the connection rights of implicit units in multilayer networks.

Currently, BP algorithms are mainly used in the following areas:

a) Function Approximation: training a network with input vectors and corresponding output vectors to approximate a function.

b) Pattern recognition: using a to-be-determined output vector to relate it to the input vector.

c) Classification: classify the input vectors in the appropriate way defined by them.

d) Data compression: reduce the output vector dimension for transmission or storage. 1)

Network structure of the algorithm

The basic idea of the BP algorithm is that the corrections to the network weights and thresholds cause the error function to decrease along the gradient direction. In this network, each processing unit has a non-linear input/output relationship, and its function function is usually a Sigmoid function. [32]

The learning process of the BP network is divided into two parts: forward propagation and back propagation of error. In the first stage (forward propagation process), the input samples pass through the input layer and are processed layer by layer by the implicit layer, and the actual output value of each unit is calculated. In the second stage (back propagation process), if there is an error between the actual output of the output layer and the desired output value, the difference between the actual output and the desired output (i.e., the error) is calculated recursively layer by layer, and this error is the basis for correcting the weights of neurons in each layer. This error is the basis for correcting the weights of neurons in each layer. The process of continuously adjusting the weights is the learning and training process of the network until the error reaches the desired level, and then the network learning process ends.

In a three-layer BP network, the input vector is $X = {(x_{1}, x_{2}, \dots, x_{1}, \dots, x_{n})}^{τ}$ , the output vector of the hidden layer is $Y = {(y_{1}, y_{2}, \dots, y_{j}, \dots y_{m})}^{T}$ , the output vector of the output layer is $O = {(o_{1}, o_{2}, \dots, o_{k}, \dots o_{l})}^{T}$ , and the desired output vector is $d = {(d_{1}, d_{2}, \dots d_{k}, \dots d_{l})}^{T}$ . The weight matrix between the input section to the hidden layer is denoted by V, $V = (V_{1}, V_{2}, \dots V_{,}, \dots, V_{m})$ , where V, is the weight vector corresponding to the j th neuron of the hidden layer. The weight matrix between the hidden layer to the output layer is denoted by W, $W = (W_{1}, W_{2}, \dots W_{k}, \dots, W_{l})$ , where W_k is the weight vector corresponding to the k th neuron of the hidden layer.

The unipolar Sigmoid function is used for the action function: (17) $f (x) = \frac{1}{1 + e^{- x}}$

For the input vector X, the relevant formulae for the neuronal network are as follows:

The input net_k to the output layer node k is: (18) $n e t_{k} = \sum_{j = 0}^{m} w_{j k} y_{j}$

The actual output o_k of output layer node k is: (19) $o_{k} = f (n e t_{k})$

The input net_j of the hidden layer node j is: (20) $n e t_{j} = \sum_{t = 0}^{n} v_{i j} x_{i}$

The actual output y_j of the hidden layer node j is: (21) $y_{j} = f (n e t_{j})$

An output error exists when the actual output of the BP network is not equal to the desired output, defined as: (22) $E = \frac{1}{2} \sum_{k = 1}^{l} {(d_{k} - o_{k})}^{2}$

Expand the above output error definition equation to the hidden layer: (23) $\begin{array}{rcl} E & = & \frac{1}{2} \sum_{k = 1}^{l} {(d_{k} - o_{k})}^{2} = \frac{1}{2} \sum_{k = 1}^{l} {(d_{k} - f (n e t_{k}))}^{2} \\ = & \frac{1}{2} \sum_{k = 1}^{l} {(d_{k} - f (\sum_{j = 0}^{m} w_{f k} y_{l}))}^{2} \end{array}$

Expand further to the input layer: (24) $\begin{array}{rcl} E & = & \frac{1}{2} \sum_{k = 1}^{l} {(d_{k} - f (\sum_{j = 0}^{m} w_{j k} f (n e t_{j})))}^{2} \\ = & \frac{1}{2} \sum_{k = 1}^{l} (d_{k} - f {(\sum_{j = 0}^{m} w_{j k} f (\sum_{i = 0}^{n} v_{i j} x_{j}))}^{2} \end{array}$

It can be seen from the formula that the output error of the network is a function of the weights w_jk, v_ij of each layer, and therefore adjusting the weights changes the error E.

Obviously, the adjustment of the weights is precisely to make the error decreasing, so the adjustment of the weights should be made proportional to the decrease of the error, i.e.: (25) $Δ w_{j k} = - η \frac{\partial E}{\partial w_{j k}}$ (26) $Δ v_{i j} = - η \frac{\partial E}{\partial v_{i j}}$

Only the expression for the weight adjustment of the three-layer BP algorithm is given below: (27) $Δ w_{j k} = η δ_{k}^{o} y_{j} = η (d_{k} - o_{k}) o_{k} (1 - o_{k}) y_{j}$ (28) $Δ v_{j j} = η δ_{j}^{y} x_{i} = η (\sum_{k = 1}' δ_{k}^{o} w_{j k}) y_{j} (1 - y_{j}) x_{i}$

where the output layer error is: (29) $δ_{k}^{o} = (d_{k} - o_{k}) o_{k} (1 - o_{k})$

The hidden layer error is: (30) $δ_{j}^{y} = (\sum_{k = 1}' δ_{k}^{o} w_{j k}) y_{j} (1 - y_{j})$

From the two equations, it can be seen that the learning rate η, the output error δ of this layer and the input vector X (or Y) of this layer determine the adjustment of the weights. 2)

Learning process of BP algorithm

The specific learning process of the BP algorithm is divided into the following 10 steps: (1)

Initialisation, i.e., assign random numbers to the weights of each node in the hidden and output layers, set the learning rate $η \in (0, 1)$ , and the expected error.

(2)

Provide the training pattern, i.e., select a training set sample from the set of training patterns and feed its output pattern and desired output into the network.

(3)

Calculate the input net_j of the hidden layer node from the input layer node x_i, the weights v_ij of the input layer and the hidden layer, using the formula, and calculate its output value $y_{J} = f (n e t_{j})$ .

(4)

From the output of the hidden layer node y_J, the weights of the hidden layer and the output layer w_jk, calculate the input of the output layer node net_k using the formula and calculate its output value $o_{k} = f (n e t_{k})$ .

(5)

Calculate the network output error.

(6)

Calculate the output layer error $δ_{k}^{o}$ and the hidden layer error $δ_{j}^{y}$ using Eq.

(7)

Adjust the weights using the formula.

(8)

Return to step (2) by taking the next sample in turn until every training pattern in the training pattern set has been learnt.

(9)

Check if the total error of the network is less than the expected error. If yes, proceed to the next step. Otherwise, return to step (2).

(10)

Record the weights and end the training.

2.2.2

Design of the neural network prediction model

1)

Design of input and output layers. The input and output layers are primarily responsible for receiving input data and producing final processing results. In general, the number of inputs in the input layer of the network should be equal to the number of inputs of the processed problem, and the number of neurons in the output layer should be equal to the number of outputs of the processed problem. The number of neurons used in the input layer of this network model is 10, and the number of neurons used in the output layer is 1.

2)

Hidden layer design: BP neural network with a hidden layer so that it can deal with more complex problems, generally contains a hidden layer or 2 layers, the number of neurons according to the actual situation of flexible choice is the network with the highest accuracy [33]. This network model is chosen to contain one hidden layer, and the number of neurons is 15.

3)

Preprocessing of neural network data. In order to make the training of the neural network more effective and to improve the training speed of the established neural network, the input and output data of the neural network should be preprocessed first. The preprocessing methods provided by Matlab include normalization, standardization, and principal component analysis. We most often use normalisation processing. That is, the input and output data are mapped to the range [-1, 1] and then reflected to the original data range after training in a certain way. Once again, the data should be disordered and categorised. 20% of the input 104 sets of data are used as test data. 20% of the sample was used as change data. The other 60% of the group is used for training with normal input.

4)

Initialisation of the neural network. Matlab initialises the network through the initff function, which can automatically call the initialisation function initff while creating the network object through newff to initialise the connection weights and thresholds of the neural network according to the default parameters of the network model.

5)

Neural network training. Generally, the training process of the BP neural network is achieved through the train-by-function. But generally, in practical applications, the algorithm of this BP is slow and difficult to achieve good results so that we can use a variety of improved algorithms. The traingdx algorithm (variable learning rate momentum gradient descent algorithm) is used for this network model.

3

Results and discussion

3.1

Research on students’ mental health assessment based on fuzzy comprehensive judgement model

3.1.1

Construction of the Mental Health Assessment Indicator System for Higher Education Students

The author A higher vocational colleges and universities as an empirical research object, here from the perspective of college students’ mental health to establish a questionnaire, the investigation of factors in the specific three large indicators continue to be divided into the questionnaire’s 18 questions, and then after the evaluation of the expert scoring, remove 3, a total of 15 questions to analyse, and further refine the 10 indicators affecting the mental health of education in colleges and universities. Therefore, this paper selects the following indicators to constitute the evaluation index system of factors affecting mental health education in a higher vocational college, including three first-level indicators and 10 second-level indicators. The evaluation index system for college students’ mental health education in higher vocational colleges is shown in Table 1.

Table 1.

Education evaluation index system of college students’ mental health education

Primary indicator	Secondary indicator
Student factor(A₁)	Student interpersonal communication problem(A₁₁)
	Students learn to be stressed(A₁₂)
	Students are not responsible for college students(A₁₃)
	Students are influenced by love(A₁₄)
	The students have little future(A₁₅)
	Job pressure(A₁₆)
School factor(B₂)	The school lacks certain mental health training(B₂₁)
	The impact of the school environment(B₂₂)
	The influence of university regulations(B₂₃)
Teacher factor(C₁)	The teacher’s care for the students(C₁₁)

3.1.2

Determination of indicator weights

The weights of the indicators at each level are shown in Table 2.

Table 2.

The weights of each level

Primary indicator	weighting	Secondary indicator	weighting
A₁	0.625	A₁₁	0.155
		A₁₂	0.085
		A₁₃	0.125
		A₁₄	0.155
		A₁₅	0.057
		A₁₆	0.048
B₂	0.276	B₂₁	0.125
		B₂₂	0.100
		B₂₃	0.051
C₁	0.099	C₁₁	0.099

This results in a vector of tier 1 indicator weights:

$W = {0.625, 0.276, 0.099}$

The vector of secondary indicator weights is:

$W_{1} = {0.155, 0.085, 0.125, 0.155, 0.057, 0.048}$

$W_{2} = {0.125, 0.100, 0.051}$

$W_{3} = {0.099}$

3.1.3

Establishment of a rubric

The set of rubrics is established as $V = {v_{1}, v_{2}, v_{3}, v_{4}}$ , i.e., let’s say there are 4 possible rubrics, where good mental health is represented by v₁, normal mental health is represented by v₂, mildly abnormal mental health is represented by v₃, and severely abnormal mental health is represented by v₄, which can be directly translated as V={good mental health, normal mental health, mildly abnormal mental health, severely abnormal mental health}.

3.1.4

Fuzzy affiliation evaluation matrix construction

The main purpose of the univariate fuzzy evaluation is to obtain the fuzzy membership evaluation matrix. For example, for a student’s mental health status, 70% of the teachers thought it was “good mental health”, 15% of the judges thought it was “normal mental health”, 10% of the judges thought it was “mild abnormal mental health”, and 5% of the judges thought it was “mild abnormal mental health”, then the degree of the student’s mental health status belonging to the evaluation set was R=(0.7, 0.15,0.1,0.05). Based on the results of the teacher survey evaluation, the fuzzy membership matrix for each indicator was determined. Table 3 shows the evaluation matrix of fuzzy membership.

Table 3.

Fuzzy membership evaluation matrix

Primary indicator	Secondary indicator	v₁	v₂	v₃	v₄
A₁	A₁₁	0.6	0.4	0.2	0
	A₁₂	0.6	0.3	0.1	0
	A₁₃	0.5	0.3	0.1	0.1
	A₁₄	0.6	0.2	0.1	0.1
	A₁₅	0.7	0.3	0	0
	A₁₆	0.7	0.2	0.1	0
B₂	B₂₁	0.4	0.3	0.1	0.2
	B₂₂	0.6	0.2	0.2	0
	B₂₃	0.8	0.1	0.1	0
C₁	C₁₁	0.6	0.1	0.2	0.1

3.1.5

Calculation of a fuzzy integrated evaluation matrix

$S_{1} = W_{1} * R_{1} = {0.155, 0.085, 0.125, 0.155, 0.057, 0.048} * {\begin{matrix} 0.6 & 0.4 & 0.2 & 0 \\ 0.6 & 0.3 & 0.1 & 0 \\ 0.5 & 0.3 & 0.1 & 0.1 \\ 0.6 & 0.2 & 0.1 & 0.1 \\ 0.7 & 0.3 & 0 & 0 \\ 0.7 & 0.2 & 0.1 & 0 \end{matrix}} = {0.373, 0.183, 0.072, 0.028}$

$S_{2} = W_{2} * R_{2} = {0.125, 0.1, 0.051} * {\begin{matrix} \begin{matrix} 0.4 & 0.3 & 0.1 & 0.2 \end{matrix} \\ \begin{matrix} \begin{matrix} \begin{matrix} 0.6 & 0.2 \end{matrix} & 0.2 \end{matrix} & 0 \end{matrix} \\ \begin{matrix} \begin{matrix} \begin{matrix} 0.8 & 0.1 \end{matrix} & 0.1 \end{matrix} & 0 \end{matrix} \end{matrix}} = {0.151, 0.063, 0.038, 0.025}$

$S_{3} = W_{3} * R_{3} = 0.099 * {0.6, 0.1, 0.2, 0.1} = {0.059, 0.010, 0.020, 0.010}$

Based on the above results, the evaluation matrix for the level 1 indicators was calculated:

$S = W * R = {0.625, 0.276, 0.099} * {\begin{matrix} \begin{matrix} 0.373 & \begin{matrix} 0.183 & \begin{matrix} 0.072 & 0.028 \end{matrix} \end{matrix} \end{matrix} \\ \begin{matrix} 0.151 & \begin{matrix} 0.063 & \begin{matrix} 0.038 & 0.025 \end{matrix} \end{matrix} \end{matrix} \\ \begin{matrix} 0.059 & \begin{matrix} 0.010 & \begin{matrix} 0.020 & 0.010 \end{matrix} \end{matrix} \end{matrix} \end{matrix}} = {0.281, 0.132, 0.058, 0.025}$

According to the comprehensive evaluation results of the above indicator layers, the comprehensive evaluation result $S = W * R = {0.281, 0.132, 0.058, 0.025}$ is obtained, and as seen from the calculation results, the evaluation result of the mental health of college students in A higher vocational college is good.

3.2

Application of predictive modelling

3.2.1

Simulation tests

The established neural network model is entered, and the simulation test is performed on the test data. The output of the network was inverted and normalized, and the results obtained from the model were compared with the ideal output. The simulation test is shown in Figure 2. Through the simulation test on several groups of data, it can be seen that the error between the prediction results of each group and the real value is very small, the degree of fitting to the input data basically meets the required requirements, and the model is able to make a good prediction of the mental health status of college students in higher vocational colleges and universities.

3.2.2

Evaluation of model performance

The comparison of the error before and after optimisation is shown in Figure 3. As can be seen from the figure, the comparison of the error before and after optimisation using the genetic algorithm reveals that the optimised neural network prediction model is better than the pre-optimised neural network, and that the optimised neural network is able to fit the data more quickly.

In this paper, we mapped the quantitative data of 14 influencing factors affecting the mental health status of college students in higher vocational colleges and universities (policy employment stressor, social employment stressor, school employment stressor, personal employment stressor, policy academic stressor, school academic stressor, personal academic stressor, policy economic stressor, school economic stressor, personal economic stressor, gender stressor, age stressor, role conflict stressor, teacher-student relationship stressor) to the corresponding mental health status and established a prediction model for some sample data as shown in Table 4. stressors, teacher-student relationship stressors) of the quantitative data were mapped to the corresponding mental health conditions to establish a prediction model for the mental health of college students in higher vocational colleges and universities, and the results of some sample data were analysed as shown in Table 4. The results show that the prediction model is able to make good predictions for students with psychological problems through the quantification of each influencing factor by students, which basically achieves the expected purpose.

Table 4.

The results of some sample Numbers are analyzed

Serial number	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1	4	4	4	2	2	4	3	4	1	1	4	1	2	1
2	3	3	3	1	2	1	2	3	1	3	3	3	4	1
3	3	4	4	1	3	3	1	1	3	1	1	1	3	2
4	3	3	4	2	2	2	4	4	4	4	3	3	1	1
5	1	1	1	4	4	2	4	4	4	1	3	2	2	4
6	1	3	2	4	2	1	1	1	1	4	2	1	3	3
7	2	1	3	4	2	2	1	1	4	1	4	4	2	3
8	2	1	3	2	1	4	3	2	1	2	3	2	3	2
9	1	2	4	2	3	1	3	2	4	4	2	4	3	4
10	1	3	3	1	2	4	4	1	3	2	3	3	1	1
11	2	3	1	2	3	3	1	1	1	4	1	2	1	2
12	3	1	4	2	3	2	2	2	2	3	1	2	4	4
13	3	4	3	3	2	2	1	1	3	2	3	1	3	1
14	2	2	2	3	2	4	4	3	4	4	4	2	2	1
15	1	3	3	2	2	2	2	4	1	1	3	3	2	3
16	1	2	4	1	2	1	3	3	2	3	4	3	2	4
17	3	1	4	3	1	1	1	3	2	1	3	1	3	3
18	3	2	1	1	3	3	4	3	2	3	1	2	4	3
19	1	1	2	3	1	2	3	2	2	3	2	2	4	4
20	2	1	2	2	3	4	4	1	3	2	1	4	4	2

The results of some of the sample data were analysed as shown in Table 5.

Table 5.

The results of some sample Numbers are analyzed

Serial number	Forecast output	Standard value	Absolute error	Relative error
1	1.797	2	-0.208	-9.15%
2	1.878	2	-0.11	-0.0675%
3	2.157	1	0.146	8.95%
4	2.002	2	0.011	0.95%
5	2.076	3	0.073	4.55%
6	1.021	1	0.021	2.1%
7	2.177	2	0.204	9.45%
8	2.11	2	0.112	5%
9	2.167	2	0.172	9.25%
10	1.839	3	-0.171	-8.4%
11	2.127	1	0.136	6.15%
12	2.731	3	-0.254	-9.4%
13	2.916	3	-0.062	-2.63%
14	2.003	2	0.007	-0.75%
15	2.779	3	-0.205	-6.63%
16	3.296	3	0.294	10.33%
17	3.26	3	0.237	8.33%
18	2.793	2	-0.22	-6.93%
19	1.078	1	0.077	8.7%
20	2.83	1	-0.18	5.43%

3.2.3

Analysis of practical applications of forecasting models

The neural network prediction model established in this paper can basically output the mental health status of college students in higher vocational colleges and universities effectively by analysing the data of the quantitative results of the main factors affecting the mental health status of college students in higher vocational colleges and universities. The prediction results are only a representation of the student’s psychological state in the recent period. Through the prediction of the psychological health status of college students in higher vocational colleges and universities, we can understand the current psychological health status of the students so that we can better carry out the work of mental health education. Through the established prediction model, the psychological health status of several college students in higher vocational colleges was predicted, and the personality questionnaire of college students in UPI was used to validate and analyse the prediction model. The results of the model were 1 for good mental health, 2 for normal mental health, 3 for mildly abnormal mental health, and 4 for severely abnormal mental health. The results of the University Personality Inventory (UPI) are represented as A for mental health problems. B for no serious mental health problems. C for neither of the above. A comparison of the predictive model with the UPI measure is shown in Table 6. The results of the data show that the prediction model of the mental health status of college students in higher vocational colleges established in this paper has a very high practical value in real life and application and can more accurately predict the mental health status of the college student population in higher vocational colleges and universities, and basically can achieve the expected results.

Table 6.

Comparison of prediction model and PI measurement

Serial number	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1	4	2	1	3	1	2	2	4	2	3	4	4	2	4
2	4	1	2	1	2	1	1	4	2	3	4	2	1	4
3	3	2	4	3	2	1	4	1	4	3	1	2	3	2
4	3	2	3	1	4	4	2	4	3	4	3	3	4	1
5	3	1	1	3	4	2	2	2	1	4	2	2	2	2
6	1	2	4	4	2	1	1	4	3	1	2	4	4	1
7	4	1	3	4	2	1	2	2	1	1	2	1	3	4
8	2	4	3	4	1	4	4	1	3	4	3	4	3	2
9	2	3	4	1	2	2	3	3	3	1	3	4	3	2
10	1	2	2	4	1	2	1	1	2	2	2	4	1	2
11	4	4	2	1	1	1	4	3	3	4	1	2	2	3

Comparison of the predictive model with the UPI measurements (II) is shown in Table 7.

Table 7.

Comparison of prediction model and PI measurement

Serial number	Model prediction	UPI score	Predictive result	UPI result
1	2.103	34	2	A
2	2.572	35	3	A
3	1.991	31	1	B
4	2.593	38	2	B
5	2.823	40	3	B
6	2.052	29	2	A
7	2.45	35	3	C
8	2.627	37	3	A
9	2.221	35	2	B
10	1.549	21	2	B
11	2.639	42	3	A

4

Research on countermeasures for mental health interventions for college students in tertiary institutions

4.1

Designing intervention strategies from the perspective of HEIs

For depressed college students, colleges and universities can adopt various intervention strategies to effectively improve their mental health.

First, colleges and universities can set up psychological support teams composed of professional psychologists, psychological counsellors, social workers, volunteers, etc., and establish mental health service centres to provide different forms of mental health services. For example, psychological counseling, psychological assessments, and treatment services. Through diversified forms of services, all-around psychological support helps to better meet the needs of different students and help depressed college students solve their psychological problems in time.

Secondly, online psychological counseling and telepsychotherapy are used to provide students with more convenient and flexible mental health services. These services can be provided by means of online video conferencing or telephone counselling, etc., so that students can easily obtain psychological support and treatment at home or elsewhere.

Thirdly, tertiary institutions can offer relevant mental health programs and activities in the classroom to enhance mental health education for students. For example, they can provide students with training in psychological knowledge and skills to help them improve their psychological quality and ability to cope with stress. In addition, colleges and universities can organise students to participate in sports, music, art and other activities, so as to provide a good environment for students to relax and release pressure, which will serve to alleviate their symptoms.

4.2

Designing intervention strategies from a teacher’s perspective

Teachers in colleges and universities should pay continuous attention to the mental health of students, pay attention to the emotional changes of students, and have timely and in-depth exchanges with them to understand their psychological state in order to truly grasp their problems and confusion, and give them the necessary help and support. Especially in the case of excessive pressure in learning and life, they should be given psychological comfort and solace to help them relieve stress. 1)

Providing personalised teaching support and guidance in classroom teaching

Teachers can pay attention to students’ learning situation and learning progress in classroom teaching, provide personalised teaching support and guidance, help students build up their self-confidence, improve their learning effect and reduce academic pressure. At the same time, they can also focus on cultivating students’ comprehensive qualities and abilities, promoting their all-round development, enhancing students’ self-confidence and self-esteem, and improving their mental health. In addition, teachers should create a favourable classroom atmosphere, establish a good teacher-student relationship with students so that students feel the love and care of the teacher, stimulate students’ interest in learning, motivate students to study diligently, and continuously improve the learning effect.

2)

Organise regular mental health class meetings to create a relaxed and positive atmosphere

Teachers, such as classroom teachers and counsellors, can hold regular mental health classes, using lighthearted and humorous ways to help students reduce stress. For example, some funny videos or cartoons can be made to exaggerate the situations in learning and life, with funny soundtracks, to help students reduce negative emotions. In addition, some interesting activities such as games, speeches and competitions can be organised so that students can build up positive emotions and optimistic attitudes in a relaxed and happy atmosphere, thus reducing negative emotions, boosting self-confidence and improving learning efficiency.

5

Conclusion

In recent years, with the rapid development of computers and intelligent technology, they have increasingly played an important role and had an increasing influence on the development of college students’ mental health education in higher vocational colleges. In this context, the article researches and discusses the innovative path of college students’ mental health education in higher vocational colleges. Through empirical testing, this article concludes:

The comprehensive evaluation result of the psychological health of college students in A higher vocational college is $S = W * R = {0.281, 0.132, 0.058, 0.025}$ , and the psychological health status of college students in A higher vocational college is good.

By evaluating the performance of the prediction model for college student’s mental health status in higher vocational colleges constructed in this paper, the results show that the predicted value is closer to the actual value. The constructed model is compared with the international UPI college student personality test for practical application, and the results show that the model is able to better predict the psychological health status of college students, and to a certain extent, it provides valuable research for the work of psychological education in colleges and universities.

Language:: English

Publication timeframe:: 1 times per year
Journal Subjects:: Life Sciences, Life Sciences, other, Mathematics, Applied Mathematics, General Mathematics, Physics, Physics, other

Journal RSS Feed

Serial number	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1	4	4	4	2	2	4	3	4	1	1	4	1	2	1
2	3	3	3	1	2	1	2	3	1	3	3	3	4	1
3	3	4	4	1	3	3	1	1	3	1	1	1	3	2
4	3	3	4	2	2	2	4	4	4	4	3	3	1	1
5	1	1	1	4	4	2	4	4	4	1	3	2	2	4
6	1	3	2	4	2	1	1	1	1	4	2	1	3	3
7	2	1	3	4	2	2	1	1	4	1	4	4	2	3
8	2	1	3	2	1	4	3	2	1	2	3	2	3	2
9	1	2	4	2	3	1	3	2	4	4	2	4	3	4
10	1	3	3	1	2	4	4	1	3	2	3	3	1	1
11	2	3	1	2	3	3	1	1	1	4	1	2	1	2
12	3	1	4	2	3	2	2	2	2	3	1	2	4	4
13	3	4	3	3	2	2	1	1	3	2	3	1	3	1
14	2	2	2	3	2	4	4	3	4	4	4	2	2	1
15	1	3	3	2	2	2	2	4	1	1	3	3	2	3
16	1	2	4	1	2	1	3	3	2	3	4	3	2	4
17	3	1	4	3	1	1	1	3	2	1	3	1	3	3
18	3	2	1	1	3	3	4	3	2	3	1	2	4	3
19	1	1	2	3	1	2	3	2	2	3	2	2	4	4
20	2	1	2	2	3	4	4	1	3	2	1	4	4	2

Serial number	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1	4	2	1	3	1	2	2	4	2	3	4	4	2	4
2	4	1	2	1	2	1	1	4	2	3	4	2	1	4
3	3	2	4	3	2	1	4	1	4	3	1	2	3	2
4	3	2	3	1	4	4	2	4	3	4	3	3	4	1
5	3	1	1	3	4	2	2	2	1	4	2	2	2	2
6	1	2	4	4	2	1	1	4	3	1	2	4	4	1
7	4	1	3	4	2	1	2	2	1	1	2	1	3	4
8	2	4	3	4	1	4	4	1	3	4	3	4	3	2
9	2	3	4	1	2	2	3	3	3	1	3	4	3	2
10	1	2	2	4	1	2	1	1	2	2	2	4	1	2
11	4	4	2	1	1	1	4	3	3	4	1	2	2	3

Serial number	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1	4	4	4	2	2	4	3	4	1	1	4	1	2	1
2	3	3	3	1	2	1	2	3	1	3	3	3	4	1
3	3	4	4	1	3	3	1	1	3	1	1	1	3	2
4	3	3	4	2	2	2	4	4	4	4	3	3	1	1
5	1	1	1	4	4	2	4	4	4	1	3	2	2	4
6	1	3	2	4	2	1	1	1	1	4	2	1	3	3
7	2	1	3	4	2	2	1	1	4	1	4	4	2	3
8	2	1	3	2	1	4	3	2	1	2	3	2	3	2
9	1	2	4	2	3	1	3	2	4	4	2	4	3	4
10	1	3	3	1	2	4	4	1	3	2	3	3	1	1
11	2	3	1	2	3	3	1	1	1	4	1	2	1	2
12	3	1	4	2	3	2	2	2	2	3	1	2	4	4
13	3	4	3	3	2	2	1	1	3	2	3	1	3	1
14	2	2	2	3	2	4	4	3	4	4	4	2	2	1
15	1	3	3	2	2	2	2	4	1	1	3	3	2	3
16	1	2	4	1	2	1	3	3	2	3	4	3	2	4
17	3	1	4	3	1	1	1	3	2	1	3	1	3	3
18	3	2	1	1	3	3	4	3	2	3	1	2	4	3
19	1	1	2	3	1	2	3	2	2	3	2	2	4	4
20	2	1	2	2	3	4	4	1	3	2	1	4	4	2

Serial number	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1	4	2	1	3	1	2	2	4	2	3	4	4	2	4
2	4	1	2	1	2	1	1	4	2	3	4	2	1	4
3	3	2	4	3	2	1	4	1	4	3	1	2	3	2
4	3	2	3	1	4	4	2	4	3	4	3	3	4	1
5	3	1	1	3	4	2	2	2	1	4	2	2	2	2
6	1	2	4	4	2	1	1	4	3	1	2	4	4	1
7	4	1	3	4	2	1	2	2	1	1	2	1	3	4
8	2	4	3	4	1	4	4	1	3	4	3	4	3	2
9	2	3	4	1	2	2	3	3	3	1	3	4	3	2
10	1	2	2	4	1	2	1	1	2	2	2	4	1	2
11	4	4	2	1	1	1	4	3	3	4	1	2	2	3

Research on the Innovative Path of Mental Health Education for College Students in Higher Vocational Colleges and Universities Aided by Computer and Intelligent Technology

Liang Xu

Published Online: Feb 05, 2025

Received: Sep 24, 2024

Accepted: Dec 29, 2024

DOI: https://doi.org/10.2478/amns-2025-0082

KeywordsFuzzy comprehensive evaluation method, BP neural network, Mental health education, Psychological prediction

© 2025 Liang Xu, published by Sciendo

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

Keywords
Fuzzy comprehensive evaluation method, BP neural network, Mental health education, Psychological prediction

Serial number	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1	4	4	4	2	2	4	3	4	1	1	4	1	2	1
2	3	3	3	1	2	1	2	3	1	3	3	3	4	1
3	3	4	4	1	3	3	1	1	3	1	1	1	3	2
4	3	3	4	2	2	2	4	4	4	4	3	3	1	1
5	1	1	1	4	4	2	4	4	4	1	3	2	2	4
6	1	3	2	4	2	1	1	1	1	4	2	1	3	3
7	2	1	3	4	2	2	1	1	4	1	4	4	2	3
8	2	1	3	2	1	4	3	2	1	2	3	2	3	2
9	1	2	4	2	3	1	3	2	4	4	2	4	3	4
10	1	3	3	1	2	4	4	1	3	2	3	3	1	1
11	2	3	1	2	3	3	1	1	1	4	1	2	1	2
12	3	1	4	2	3	2	2	2	2	3	1	2	4	4
13	3	4	3	3	2	2	1	1	3	2	3	1	3	1
14	2	2	2	3	2	4	4	3	4	4	4	2	2	1
15	1	3	3	2	2	2	2	4	1	1	3	3	2	3
16	1	2	4	1	2	1	3	3	2	3	4	3	2	4
17	3	1	4	3	1	1	1	3	2	1	3	1	3	3
18	3	2	1	1	3	3	4	3	2	3	1	2	4	3
19	1	1	2	3	1	2	3	2	2	3	2	2	4	4
20	2	1	2	2	3	4	4	1	3	2	1	4	4	2

Serial number	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1	4	2	1	3	1	2	2	4	2	3	4	4	2	4
2	4	1	2	1	2	1	1	4	2	3	4	2	1	4
3	3	2	4	3	2	1	4	1	4	3	1	2	3	2
4	3	2	3	1	4	4	2	4	3	4	3	3	4	1
5	3	1	1	3	4	2	2	2	1	4	2	2	2	2
6	1	2	4	4	2	1	1	4	3	1	2	4	4	1
7	4	1	3	4	2	1	2	2	1	1	2	1	3	4
8	2	4	3	4	1	4	4	1	3	4	3	4	3	2
9	2	3	4	1	2	2	3	3	3	1	3	4	3	2
10	1	2	2	4	1	2	1	1	2	2	2	4	1	2
11	4	4	2	1	1	1	4	3	3	4	1	2	2	3