Computer public course teaching based on improved machine learning and neural network algorithm

Abstract

With the continuous progress of the times, the development of college education is also constantly tending to enrich and diversify. In the course of curriculum setting in many colleges and universities, more and more attention is paid to the teaching of computer courses for college students. In the course of setting up and teaching, we still follow the traditional teaching mode and do not pay more attention to students’ practical practice. By observing the computer course materials selected by many colleges and universities for students, we can see that most of the textbooks still focus on arranging some exercises from the point of view that science and engineering are students, and lack the basic knowledge of the curriculum. Because the understanding and research of computer course is not deep enough, the teaching effect obtained since the course is not ideal. By studying the relevant knowledge of machine learning and some important problems in the development of neural network algorithm theory, this paper puts forward some viewpoints based on the current curriculum system in colleges and universities in order to improve the learning quality of computer courses. And hope to build a variant learning model to improve students’ interest in computer courses. The exploration and inference of some knowledge in this paper are mostly my own views, some places are not professional enough, the majority of experts and scholars can criticize and correct at will.

Keywords

Computer public course machine learning neural network algorithm teaching reform

1 Introduction

In the current social development, many colleges and universities pay more attention to the teaching of computer courses in curriculum setting, and have more detailed regulations on related learning and teaching. In general, the development direction of computer technology curriculum in colleges and universities is developing towards more and more professional and rich content, and the teaching quality of students is constantly improving. Although the current curriculum has not had a very obvious effect on students’ learning, the overall effect is still good. It can not be denied that computer education in colleges and universities in China has opened up a good situation. The relevant curriculum content also follows the development of the times, striving to create knowledge to improve social viability. With the acceleration of social development and the continuous progress of various technologies, computer curriculum teaching in colleges and universities also needs reform. The background of the development of the times is not only a good opportunity for colleges and universities to carry out curriculum reform, but also a great challenge for the future development. In the process of future development, the progress of the times will encourage colleges and universities to keep up with the pace of the development of the times and update the relevant contents of computer courses when setting up the curriculum and establishing the curriculum system structure. Colleges and universities should pay close attention to the students’ learning needs and design more reasonable teaching methods from the perspective of students.

According to the observation of the current teaching mode in colleges and universities, it can be seen that in many places, colleges and universities still adopt the traditional teaching mode when teaching computer courses. For the past teaching mode, teachers are generally the main part of the classroom, mainly responsible for the explanation of various knowledge, students are often passive or numb to accept knowledge, many times the effect is not very good. However, with the popularization and development of Internet technology, colleges and universities begin to add the Internet teaching mode to the traditional teaching mode, which greatly improves the teaching quality of computer basic courses. Besides classroom teaching, teachers can communicate with students after class by building network communication platforms such as qq group and WeChat group. In the final course examination, the selected examination method also changed from the form of offline closed paper to the way of open-book answer with computer, and the learning mode of the course became more flexible. Although the new teaching model has been greatly improved in many aspects, it has not been divorced from the track of teachers as the main body of the classroom. This teaching method will also have some more and more prominent problems in the long-term development: although some students do not actively communicate or ask questions, teachers are difficult to master the specific teaching situation.

In 2009, China further standardized the reform and development of teaching methods through the formulation of overall teaching objectives. After launching a number of talent training plans, China’s university education has also improved. Although the talent training plans and goals put forward in the process of development are various, they can not be separated from the guidance of the core goals in general, in order to train students into innovative talent services. In the process of talent training, the cultivation of innovative consciousness is essential. In order to better realize the goal of all-round development of talents, colleges and universities are carrying out computer basic curriculum education. The main problem is to train comprehensive talents with computer thinking and comprehensive view of problems for national development. In many cases, the cultivation of high-end talents in our country is cultivated in the undergraduate and graduate education stages of college students. The high-quality talents needed in the development and construction of the country are cultivated through undergraduate and graduate education in colleges and universities. If colleges and universities can always pay attention to the needs of national and social development and enrich the content of talent training plan, they can make greater contributions to social development and the promotion of national comprehensive national strength.

2 Related work

In the writing of this article, the author mainly refers to and draws lessons from ten theoretical documents, and then introduces it. Literature [1] mainly analyzes the content setting of computer network course, and points out that the current setting of computer network course is mainly influenced by the theory of reductionism. Many knowledge content and practice content are separated from the concrete reality situation. In the design of teaching methods and teaching models, the teaching contents involved will still be affected by the reductive method, and many times lack of innovation according to the needs of the development of the times, so in the specific teaching activities, Learning efficiency and teaching quality are not very good. Moreover, many textbooks about computer basic courses are not very compatible with the current teaching practice, so it is very difficult to find a suitable textbook in many textbooks. Document [2] mainly describes the computer course as a course to teach students’ operational skills, ignoring the after-class practice design of the textbook. It is not enough for students to acquire knowledge only in theory, especially for courses with high practical ability such as computer, and the lack of practice will also greatly reduce the final learning effect. Many colleges and universities have a large proportion of book content in the setting of computer curriculum content, so in the few and limited classroom teaching time, After completing the teaching plan, there is no more time for students to practice in class, which greatly affects the students’ final learning results. In the process of design and development with many teaching methods, we still follow the traditional teaching mode. On the one hand, the teaching materials are not changed and enriched according to the actual needs. In the process of teaching activities, although teachers play the main role of educational dissemination, the teaching effect is not ideal because of lack of interaction with students [3]. Main point of view put forward by the literature is that the development of the present era is constantly improving, colleges and universities are also constantly absorbing experience, introducing some advanced technology, and improving their own development advantages. With the continuous development of information technology and network technology, the reform of teaching mode is also the result of development. In order to have a better development prospect in the future, colleges and universities build a network learning platform with the help of computer technology. Constantly enrich and improve the quality of teaching content. At present, many colleges and universities have set up a relatively complete information technology platform, browsing related web pages and entering related platforms can understand the different information of school development from many angles. And in many learning resources sharing platform can obtain the first-hand information of university education resources update in time. In this way, we can promote the popularity of the school, publicize the education of the school, and form more effective and convenient communication and communication among the colleges and universities, and realize the sharing of various advanced resources. With the development of the new era, the teaching management and the implementation of teaching activities in colleges and universities have become more high quality and high efficiency.

Document [4] mainly from the point of view of the development of network technology, this paper summarizes the changes made by teachers and students in teaching and learning. At the present stage of development, many colleges and universities also realize the importance and change of the development of network technology to the development of the times. Therefore, the construction of network platform and the research of network teaching method and teaching mode are also deepening. For the construction of network platform, many colleges and universities transform cognition into practice, constantly explore new network teaching methods and develop corresponding courses on network platform. This is also a more important direction in the teaching development of colleges and universities. In this paper, through the influence of network technology on the development of teaching, this paper summarizes the models and design methods put forward by colleges and universities in curriculum development, and discusses these models and methods. Document [5] mainly evaluates the means of knowledge dissemination developed by colleges and universities for network platform courses in teaching practice. The author’s view is that video as a new and more convenient means of knowledge dissemination. It does not play a role in demonstrating the course to people, or simply showing people a particular problem. Video is a new teaching method, which can stimulate and promote people’s learning in a wider range. Some knowledge can be spread through video, which can slowly affect people’s knowledge structure in the process of watching for a long time. Promote the improvement of people’s ideological realm. Literature [6] analyzes the current situation of the development of various countries in the world, and many schools have built their own open teaching projects through the network platform. In the world, each country has become the understanding of open resource sharing, and has formed a joint teaching resource exchange organization based on the state. Literature [7] summarizes the development and operation of open teaching resources at present. The speed of development and renewal of the times is very fast, so the development of open resources is constantly changing. The initial operation mode and method only show some teaching resources teaching results on the platform. After continuous development, it has been transformed into a professional network learning platform for many people who like to learn in society. There is a richer curriculum structure on the platform, as well as quite high-quality teaching resources. Literature [8] mainly discusses the relevant literature and practice of variant exercises abroad. Through the study of variant exercises, many medical problems can be solved and some special medical phenomena can also be measured and explained in detail. In the field of teaching, the research is relatively small, and the proportion of practice in the teaching process is relatively small. In the process of teaching, the concept of variant practice has not been formed.

literature [9] traces the discussion of variant exercises back to the concept of variant put forward by early famous psychologist Witton, but by further reading his related works, we can see that his understanding of variant exercises is not comprehensive enough. The statement put forward in this paper is not accurate enough. His main reference is still around change, and the discussion of variant exercises is not comprehensive and accurate. In his research, the variant is understood as a method that can consciously control people’s learning ability. Literature [10] summarizes the development and important role of education. In the process of social development and the continuous progress of the times, education has always been one of the important driving forces and sources to promote people’s cultural progress and social development. Education is also an important product of the development of the times [11]. In other words, the emphasis of education will change according to the development of the times, and the development of education can also promote the progress of the times [12, 13]. In the process of the development of higher education, colleges and universities not only follow the pace of the development of the times, constantly reform their own education system, can not only stay in the current stage of development, but also focus on the longer term. Let the progress of education continue to promote social development [14 –16]. The level of higher education mainly depends on the quality of graduate education, and the quality of education depends on the specific curriculum structure and the choice of teaching methods. This paper mainly analyzes the learning quality and effect of computer courses offered by colleges and universities for graduate students who are not computer majors, probes into the practice of relevant computer basic courses, analyzes the learning ability of graduate students in the process of learning, and the learning characteristics of scientific and technological research ability and final thesis writing process [17]. After further investigating the situation of students in different majors studying computer basic courses, combining with the consideration of many factors, a more reasonable curriculum system is designed for students from a scientific point of view. Finally, the teaching should start from the students, the curriculum should meet the students’ practical learning needs, grasp the rules of students’ learning, and improve the teaching quality in the future [18 –20].

3 University public computer network course

This paper studies the setting of computer network, which is a course in computer technology, not in the background of computer network. At present, the teaching materials of computer courses can be chosen casually, so they are more flexible in the selection of teaching materials [21 –23]. Although there is no clear outline to explain the relevant content of the course, there are still two major aspects that can be used as a reference, on the one hand, the content of computer technology, on the other hand, the content of communication technology [24, 25]. In the course of teaching in colleges and universities, the basic course of computer network is a part of the computer curriculum module and a more important content. For how to set the content of computer courses, some colleges and universities usually choose the teaching materials of basic courses to introduce them, and some colleges and universities follow the trend of curriculum reform. The computer network is taught independently as a subject, and students can learn voluntarily. Some scholars have explained the importance of computer basic courses according to the needs of current social development, and linked the relevant contents with the future development and social development of students, which is considered to be a compulsory subject in the stage of higher education. Specific teaching objectives are as follows:

Variant learning: This method is mainly to change some concepts and rules to make students better understand and learn the relevant contents without changing the main teaching conditions.

Support strategy: This strategy is mainly a series of regulations on the learning process that students can complete high quality learning in a short time in order to improve the harvest and effect of learning. If you want to add something to assist these requirements, you can make it a support policy. The content of the support strategy includes two levels, one is not to explain and describe the process of problem solving too much, which means that the principle of problem solving can be widely used and operable. The other is to provide specific solutions to the problems needed in the learning process.

The main purpose of teaching design is to provide students with better learning conditions. Under the guidance of the syllabus, a subsystem should be designed to help students to learn. The process of design is also an innovative process, which is based on the existing teaching situation to create new things that are helpful to the concrete teaching practice. Under the guidance of certain teaching ideas, the pan-technical view in the design combines multimedia technology and network technology to develop new teaching methods to improve students’ learning ability and teachers’ teaching skills. Five Results of Gagne’s Learning as show in Fig. 1.

Fig.1

Five results of Gagne’s Learning.

(1) Classification of learning outcomes

Gagne divides the learning results into five properties, as shown below:

In his eyes, the function of verbal information is the ability to state the form of the topic in the process of proposition, to analyze the topic or to put forward some ideas. It can make a specific analysis before people solve the problem and provide good help for people to solve the problem. Wisdom skills are the ability to master the living rules and activities of people after adapting to their living and learning environment. This ability is finally manifested by cognitive rules, and the learning of this ability is also the main part of school rules; Cognitive strategy is the ability of students to regulate their learning activities by controlling their own behavior and thinking ability in the process of learning. This ability is a test of learners’ special ability in the process of learning and thinking. It can be used to regulate and control the formation of some concepts and the application of rules. The content is not specific to the external content and has universal application performance; In the process of completing a certain action, the human body controls the muscles of its body and the coordination of its movements to complete the whole action; Attitude is the psychological state of students in the process of learning. Attitude includes both the learner’s personal emotional component and the learner’s cognitive ability and the consequences of behavior. Autonomous work pattern for learning tasks as show in Fig. 2.

Fig.2

Autonomous work pattern for learning tasks.

(2) Psychological model of autonomous learning

From the above picture, students have certain dynamic and changeable characteristics when they complete the selection of learning tasks. If the task is very difficult to perform, students will choose a slightly less difficult learning task to learn or stop the completion of the task directly. If the learning process has completed the lower difficulty learning task, it can not solve the more difficult learning task, can use support strategies to help learners complete the learning and task. In the writing of this paper, the author advocates the first learning mode, students should consciously choose their own learning tasks according to their own situation in the process of learning. After completing a stage of learning, the learner should make a brief evaluation of the completion of the learning task in this stage, and decide whether to carry out the next stage of learning or not.

In this analysis, the main purpose is to see the relationship between the various courses set up in the computer major, and to analyze the relationship between curriculum learning and employment, using the method of association rule mining. Correlation between computer courses as show in Table 1.

Table 1

Correlation between computer courses

Order	Rules	Confidence levels
1	The Digital Logic and Digital (2-1) = A Computer Networks (3-1)^∧ to Object Systems Analysis and Design (3-2) = A11	1
2	C Language (1-1) = A Computer Networks (3-1) = A10 = = A Analysis and Design to Object Systems (3-2) = A10	1
3	The Digital Logic and Digital (2-1) = A Computer Operating System Principles (3-1) = A10 = = A Against Image System Analysis and Design (3-2) = A10	1
4	Digital Logic and Digital (2-1) = A Computer Operating System Principles (3-1) = A10 = = A Computer Networks (3-1) = A10	1
5	Digital Logic and Digital (2-1) = A Computer Operating System Principles (3-1) = A10 = = A Computer Network (3-1) = A Object-oriented System Analysis and Design (3-2) = A A10	1
6	Principles of Computer Operating Systems (3-1) = A Computer Networks (3-1) = A14 = = A object systems Analysis and Design (3-2) = A13	0.93
7	Introduction to Software Engineering (1)2 = C Compiler Principle (2-2) = D12 = = C Windows Programming (3-1) = C11	0.92
8	Data Structure (2-2) = B Algorithm Analysis (3-1) = A11 = = B Engineering Training (3-2) = A11 = = B B10	0.91
9	Senior Mathematics (1-2) = C Computer Operating System Principles (3-1) = C13 = = C Software Engineering Graduation Design (4-2) = Cll	0.85
10	Higher mathematics Al (1-1) = C20 = >higher mathematics A2(1-2)B = 19	0.85

It can be seen from the data in the observation table that if students do not study advanced mathematics courses and computer courses seriously, the students’ graduation design scores will be affected to a certain extent, and if students do not study computer courses seriously, Then the computer operating system will be unfamiliar, and in the future internship employment, the learning effect of these two courses is directly related to the efficiency and quality of employment. If we do not master the basic knowledge of computer courses in the course of freshmen and sophomores, then the junior will be affected when carrying out practical operation and further study. Through the data in the table, we can also see that many college courses are related to each other. If some courses are not well studied, they will affect the study of other courses. Even related to their graduation project and future life and work. In the current teaching system of colleges and universities, the assessment of students in schools is mainly based on the calculation of comprehensive credits for final studies. Schools can carry out more regular management and practical teaching guidance according to the calculation rules. Links between courses and internships as show in Table 2.

Table 2

Links between courses and internships

Order	Rules	Confidence levels
1	Discrete mathematics (1-2) = C principles of computer operating system (first 1) = C9 = = C practice = C9 = = C N9	1
2	Linear Algebra (1-2) Requirements Engineering (3-1) = B9 = = >Practice = Y9	1
3	To Object System Analysis and Design (3-2) = B Software Engineering Graduate Design (4-2) = B 9 = = = B Practice = B 9 = = = B Y	1
4	Training of Computer Basic Skills (1-1) = A Probability Theory and Mathematical Statistics (2-1) = B8 = = = A Practice = B8 = = = A Y8	1
5	Linear Algebra (1-2) Analysis and Design to Object Systems (3-2) = A8 = >Practice = Y8	1
6	Probability and Mathematical Statistics (2-1) = B Computer Composition Principle (2-2) = A8 = = B Practice = A8 = = B Y8	1
7	Probability and Mathematical Statistics (2-1) = B Software Project Management (2-2) = A8 = = = B Practice = A8 = = = B Y8	1
8	Compilation Principle (2-2) Program Design (2-2) = B8 = >Practice = Y8	1
9	Software Project Management (2-2) = A Algorithm Analysis (3-1) = B8 = = A Practice = B8 = = A Y8	1
10	Training in Basic Computer Skills (1-1) = A Software Testing Technology (3-2) = A11 = = = A Practice = A11 = = = = A Y10	0.91

According to the above data, we can infer that the course scores of discrete mathematics and computer operating system, if not very good, will affect the situation of senior students choosing practice. Some related courses around computer operation and discrete mathematics will be helpful to senior practice. Links between courses and offer as show in Table 3. Links between courses and high salaries as show in Table 4.

Table 3

Links between courses and offer

Order	Rules	Confidence confidence Degree
1	Computer Foundation (1-1) = A Software Project Management (2-2) = A12 = = = A Offer = Y12	1
2	Principle and Application of C = A Database (3-1) = All = C = A Offer = Yll	1
3	Computer Foundation (1-1) = A Software Testing Technology (3-2) = All = = = A Offer = Yll	1
4	The Data Structure (2-1) = E = All = = E Offer = Yll to Object Programming (2-2); and	1
5	Principle of Computer Operating System (3-1) = A Analysis and Design of Object-oriented System (3-2) = A AOffer = A Yl 1	0.93
6	probability theory and mathematical statistics (2 + 1) = B10 = = C Offer = Y10 of higher mathematics	0.92
7	Computer Foundation (1-1) = A Program Integrated Design (2-1) = B10 = = = A OfFer = Y10	0.91
8	Probability and Mathematical Statistics (2-1) = B Software Project Management (2-2) = B10 = = = B Offer = Y10	0.91
9	Digital Logic and Digital (2-1) = B Compilation Principle (2-2) = A10 = = = B OfFer = Y10	0.85
10	Principles of Compilation (2-2) = A of Computer Operating Systems (3-1) = A10 = = A OfFer = Y10	0.83

Table 4

Links between courses and high salaries

Order	Rules	Confidence levels
1	Program Design (2-2) = B20 = >Higher Pay = G20	1
2	Graduation Design (4-2) = B17 = = >Higher Pay = G17	1
3	Introduction to Software Engineering (1-2) = A14 = >Higher Pay = G	1
4	Software Project Management (2-2) = A14 = >High Pay = G14	1
5	Integrated Computer Operating System Design (3-2) = C14 = >Higher Pay = G14	1
6	Programming C (1-2) = A13 = = >high-paying = G13	1
7	Computer Foundation (1-1) = A13 = >Higher Pay = G13	1
8	Software testing techniques (first 2) = A13 = >high-paying = G13	1
9	Software Development Experiment (4-1) = A13 = >Higher Pay = G13	1
10	To object programming (2-2) = A16 = >high pay = G16	1

According to the data in the table, the study of computer courses and mathematics courses is not only related to the graduation design of students, but also helpful for students to find internships in senior years. It not only improves their comprehensive skills, but also determines their salary level after formal work.

4 Neural networks

The neural network is composed of neurons, and the processing unit includes several small processors. The processor can not only acquire knowledge by using neurons in the process of learning, but also store the existing knowledge in the neural network and extract it when needed. By analyzing the important role of artificial neural network, we can see that it can be widely used in some aspects of target recognition, function application, automatic control and so on in the future. Artificial neuron model as show in Fig. 3.

Fig.3

Artificial neuron model.

(1) artificial neuron model

The specific model is shown below:

Artificial neurons are nonlinear units, and the relationship between input and output can be expressed as: □ $\begin{matrix} I_{t} = \sum_{j = 1}^{n} w_{ji} \cdot x_{j} + s_{i} - θ_{i} \\ u_{i} = f (I_{i}) \\ y_{i} = g (u_{i}) = h (I_{i}) \end{matrix}$ (1)

The nonlinear characteristics of neurons can be expressed as:

1 Step function: $h (x_{i}) = f (x_{i}) 0 = U (x_{i}) = {\begin{matrix} 1, x_{i} > 0 \\ 0, x_{i} ⩽ 0 \end{matrix}$ (2)

Subsection 2: $f (x) = {\begin{matrix} 1, x ⩾ x_{0} \\ ax + b, x_{1} ⩽ x ⩽ x_{0} \\ 0, x < x_{1} \end{matrix}$ (3)

S igmoid type: $f (x) = \frac{1}{1 + exp (- x)}$ (4)

(2) BP neural networks

Calculation of output and hidden layers: □□ $δ_{k} = (T_{k 1} - O_{k}) \cdot O_{k} \cdot (1 - O_{k}) k \in {1, 2 . \dots, m}$ (5)

Calculation of hidden and input layers; □□ $δ_{j}^{l} = o_{k}^{l} \cdot (1 - o_{k}^{'}) \cdot \sum_{k - 1}^{m} δ_{k}^{l + 1} \cdot w_{jk}^{l + 1}$ (6)

After calculating the corresponding value, the values of the output layer and the hidden layer are corrected to obtain the error: $\begin{matrix} W_{jk}^{'} (t + 1) = W_{jk}^{'} (t) + α \cdot δ_{k}^{'} \cdot Y_{j}^{l - 1} \\ θ_{k}^{'} (t + 1) = θ_{k}^{'} (t) + β \cdot δ_{k}^{'} \end{matrix}$ (7)

And the following formula: □ $\begin{matrix} W_{ij}^{'} (t + 1) = W_{ij}^{'} (t) + α \cdot δ_{j}^{'} \cdot Y_{i}^{l - 1} \\ θ_{j}^{'} (t + 1) = θ_{j}^{'} (t) + β \cdot δ_{j}^{'} \end{matrix}$ (8)

(3)fuzzy neural network model

At present, there are three kinds of fuzzy neural models: □

Fuzzy neural described by fuzzy rules: According to relevant theories, the representation of fuzzy neurons can be written as: $R_{i} = A_{1}^{i} \times A_{2}^{i} \times \dots \times A_{n}^{i} \times B_{i}$ (9) Given some value, you can show it specifically; $Y_{i} = A_{1} \circ (A_{2} \circ \dots (A_{n} \circ R_{i}) \dots))$ (10)

Fuzzy neuron with real input: A mathematical representation of a fuzzy neuron can be written as: □□ $μ (x_{1}, x_{2}, \dots, x_{n}) = μ_{1} (x_{1}) ⊙ μ_{2} (x_{2}) ⊙ \dots ⊙ μ_{n} (x_{n})$ (11)

Fuzzy neuron algorithm as show in Fig. 4. Fuzzy neurons can be W mathematically described as:

Fig.4

Fuzzy neuron algorithm analysis.

$\begin{matrix} Y = x_{1} ⊙ x_{2} ⊙ \dots ⊙ x_{n} \\ x_{1}^{'} = G_{i} (x_{i}), i = 1, 2, \dots, n \end{matrix}$ (12)

Improved fuzzy neural network: As people’s understanding continues to deepen, the shortcomings of this calculation method are also constantly displayed. In order to avoid errors, further verification is needed, that is, □ $Δ w_{ij} (t + 1) = {aw}_{ij} (t) + η δ_{i} (t + 1) O_{i} (t + 1) 0 ⩽ a ⩽ 1$ (13)

The judging conditions of the training momentum term are as follows: $a = {\begin{matrix} 0 E (t) > E (t - 1) \times 1.04 \\ 0.85 E (t) < E (t - 1) \\ a Other \end{matrix}$ (14)

5 Machine learning

Machine learning plays a great role in the development and research of artificial intelligence. The development of artificial intelligence system needs to combine the knowledge of many fields and industries to perfect its own system structure, in order to better promote the upgrading of artificial intelligence system, The research and analysis of knowledge in many fields contained in machine learning can provide a good help.

The field of machine learning mainly analyzes the existing data and studies how to analyze the laws and connections between the data from the known database. By using the law of discovery to predict a future index or a feature of the classification and clustering. So far, the image semantic classification methods developed on the basis of machine learning mainly include supervised learning and unsupervised learning. The method of supervised learning can be used in a learning system that has been classified. The training set in supervised learning includes two parts: output and input, that is, the characteristics and objectives that people usually call. All goals in the training set are marked by people. The training set in the unsupervised learning method is no longer annotated artificially. When classifying the clustered data in the classification, the unsupervised learning algorithm is often used.

Naive Bayes, the main principle of Bayesian classifier is to calculate the prior probability of a certain category by Bayesian formula, and then calculate the probability sum and conditional probability of its occurrence in the category in which the category is located. The value is posterior probability. The formula is as follows: $P (C_{i} | X) > P (C_{j} | X) 1 < \exists < = m,, j \neq I$ (15)

The probability of each category can be calculated using the following formula: $P (C_{i} | X) = P (C_{i} | x_{1} \dots \dots x_{n}) = \frac{P (C_{i}) + P (x_{1} . . . . . . x_{n} | C_{I})}{P (x_{1} . . . . . . . x_{n})}$ (16)

SVM classification method is a method that people summarize and develop through long-term practice and research. In data analysis and statistics, this classification method has deep and strong data system and theoretical support. Many questions and discussion criteria about data analysis are based on statistics, so the results of analysis and evaluation are more statistically significant. In the process of data analysis, because there are more data, especially when analyzing a large number of data, it is necessary to choose the classification method of risk and small error as far as possible to carry out statistics and analysis. This method solves people’s troubles to a great extent. The formulas for common linear kernel functions are as follows. $K (x_{i}, x_{j}) = x_{i} x_{j}$ (17)

The formulas for common polynomial kernel functions are as follows. $K (x_{i}, x_{j}) = (γ x_{i}^{T} x_{j} + r)^{d}, γ > 0$ (18)

The formula of radial basis kernel function is as follows. $K (x_{i}, x_{j}) = exp (- γ {| x_{i} - x_{j} |}^{2}), γ > 0$ (19)

In the normal process of data analysis, if it is a simple data function model, we can use more commonly used analytical methods. However, if some inseparable data models and some data with less obvious features are encountered, we must use the advantages of kernel function to analyze and disassemble it. Using kernel function to analyze complex data can save the time of analyzing data to a great extent and improve the effect of data analysis. $x^{2} (t, c) \frac{N \times (AD - CB)^{2}}{(A + C) \times (B + D) \times (A + B) \times (C + D)}$ (20)

RBF formula of the kernel function is as follows. $k (x, y) = \exp (- γ {∥ x - y ∥}^{2}), (γ > 0)$ (21)

If some nonlinear data is encountered, it can be projected into the space of high dimension for observation, and the classification problem of sample X can be solved by establishing plane analysis. $w • φ (x) + b = 0$ (22)

Table 5

Computer public basic course knowledge module and hours schedule

Serial number	Knowledge module	Hours of study
Module 1	Introduction	2
Module II	Resource retrieval and management	8
Module 3	Writing specification and layout	12
Module 4	Paper defense and presentation techniques	11
Module V	Data collection, processing and statistical analysis in paper writing	10
Module 6	Paper testing	5

Table 6

The Difference between stage assessment method and traditional assessment method

Assessment	Traditional assessment methods	Stage assessment
Assessment objectives	Pay attention to the grasp of basic knowledge and theory, ignore the assessment of ability and professional quality	Flexible application of knowledge, skills, operations and thinking
Form of assessment	It is mainly based on the final examination, a test paper fixed results, a single emphasis on the results of the examination	Examination will be based on the modular features of the course content knowledge points At the end of the semester, each student’s grades at different stages are calculated proportionally, emphasizing the process of each stage of study
Learning process	The teaching mode is single, and the teaching process centered on teachers	The teaching mode is diversified, student-oriented, teacher-assisted teaching process
Learning motivation	Most students are in a passive learning state. Due to the psychology of examination-oriented examination, the study of the course is mostly concentrated in the surprise stage before the final examination	Students are expected to pass each stage of examination to get the ideal salty achievement, because the song avoids the phenomenon of lax study and surprise examination before the final examination

The data can be used to support the vector to construct Lagrange function. Its formula is as follows: $L (w, b, z) = \frac{1}{2} (w • w) - \sum_{i = 1}^{l} a_{i} [y_{i} (w • x_{i}) + b - 1], a_{i} > = 0$ (23)

Then, select the kernel function, need to meet the following formula. $k (x, y) = φ (x) • φ (y)$ (24)

Weighted K Neighborhood Algorithm. This algorithm is calculated by selecting the mean value of the previous k approximate adjacent points. In the process of calculation, we can also see that the algorithm is more flexible than the ordinary adjacent algorithm. A more accurate location result of data distribution can be obtained by selecting the appropriate k value. Computer public basic course knowledge module and hours schedule as show in Table 5; The Difference between stage assessment method and traditional assessment method as show in Table 6.

The weighted k nearest neighbor algorithm adds the related concept of weight on the basis of the k nearest neighbor algorithm, which is no longer calculated by using the average value of each point, but adds different weight values to each point. The smaller the distance from the nearest neighbor, the larger the weight. The formula for weights is as follows: $t_{i} = \frac{\frac{1}{d_{i}}}{\sum_{i = 1}^{k} \frac{1}{d_{i}}}$ (25)

In the traditional curriculum architecture, the content of many courses has been relatively old, and has been seriously disjointed with the development of society, the pertinence of professional development is not strong. Under the background of the development of the present era, combined with the theory of multiple intelligence, a modular curriculum system is formed in the evil structure of computer major courses, which is an important requirement for non-computer major graduate students to study computer basic courses. The knowledge learning of teaching materials is only a part of mastering theoretical knowledge, which needs to meet the needs of knowledge structure of computer major learning by various means.

Modularization of curriculum system design is one of the important contents of curriculum content setting, which can improve the effectiveness of curriculum framework setting. The course can be divided into different modules through the characteristics of students’ daily learning and curriculum, and students are encouraged to cross-study in order to improve their comprehensive ability. In each module system, the learning of some courses is relatively independent, but after mastering the learning of some modules, the knowledge of these modules can be combined synthetically, so as to improve their ability to use knowledge in an all-round way.

Through the examination of the curriculum, we can make a concrete analysis of the curriculum, which is an important way to directly reflect the educational and teaching activities in colleges and universities. In the construction of curriculum system, the establishment of curriculum evaluation system and assessment method is indispensable. Designing reasonable curriculum assessment methods can make scientific and effective evaluation of teachers’ teaching activities. It can provide better help and improvement for teachers’ teaching activities in the future. Through the course examination, we can make up for the defects in the curriculum, set up more curriculum contents according to the students’ needs, and fundamentally improve the students’ innovative thinking and ability.

In the current educational development, the method of periodic assessment is usually used, which is very different from the traditional way of curriculum assessment. First of all, the traditional teaching mode mainly pays attention to the role of teachers in the classroom, the main body of the classroom is teachers, through teachers constantly explain the knowledge of textbooks, students are forced or useless to receive, The teaching quality and teaching efficiency in this mode are very poor. This paper studies the influence of curriculum on the learning of non-computer majors, combines the specific ways of modularization, and synthesizes the fruit stage assessment of the final teaching results of the course. This assessment method can provide an orderly and efficient environment for classroom construction, stimulate students’ enthusiasm and create a relaxed classroom atmosphere.

6 Conclusions

This paper studies from many aspects, the learning situation and curriculum development of computer courses set up by non-computer majors in colleges and universities in China, through the concrete analysis of the practical significance and learning effect of the courses. The concept of curriculum design under the new situation is put forward. Because this paper mainly from the theoretical point of view to the future development of the proposal, has not been strongly supported by concrete practice. Therefore, in the later research, the author will explore the variant practice design of college students’ network course more deeply, so as to improve the design effect of computer network course. With the continuous development of society, many new theoretical and practical results are constantly introduced, and the content of realizing the excellent integration of computer network courses and variant exercises will be further pushed to practice.

Footnotes

Acknowledgments

Higher Education Research Fund Project of Xi’an Aeronautical University in 2019 “The Research and Practice of the Classroom Teaching Model of the Course ‘Fundametals of Computer Culture”’(2019GJ1008).

References

Kerber

R.G.

, Using a generalization hierarchy to learn fromexamples, Proceedings of the Fifth International Conference on Machine Learning 22(12) (2014), 1–7.

Zhang

and Zhao

, Transfer Effects of Learning through Two Varied Worked Examples on Word-problem Solving, Acta Psychologica Sinica 12(2) (2008), 6–12.

Paas

F.G.W.C.

and Van Merri?nboer

J.J.G.

, Variability of worked examples and transfer of geometrical problem-solving skills: A cognitive-load approach, Journal of Educational Psychology 86(1) (1994), 122–130.

Catrambone

, Predicting What People Learn From Examples, Georgia Inst Of Tech Atlanta School Of Psychology 12 (1995), 43–47.

Quilici

J.L.

and Mayer

R.E.

, Role of examples in how students learn to categorize statistics word problems, Journal of Educational Psychology 88(1) (1996), 144.

Yum

, How coherent texts with varied examples help students learn about statistics: a test of text processing models, University of Texas at Austin 23 (1998), 12–19.

Ejdys

and G6ra

, The more we learn the less we know? On inductive learning fromexamples, Fundamenta Informaticae 39(4) (1999), 359–374.

Schilling

M.A.

, Vidal

, Ployhart

R.E.

, et al., Learning by doing something else: Variation, relatedness, and the learning curve, Management Science 49(1) (2003), 39–56.

Van Merri?nboer

J.J.G.

, Kester

and Paas

, Teaching complex rather than simple tasks: Balancing intrinsic and germane load to enhance transfer of learning, Applied Cognitive Psychology 20(3) (2006), 343–352.

10.

Kaminski

, Sloutsky

and Heckler

, Concrete examples don’t help students learn math, study finds, Retrieved September 15(10) (2008), 20–23.

11.

Lsotani

, Adams

, Mayer

R.E.

, et al., Can erroneous examples help middle-school students learn decimals? Towards Ubiquitous Learning, Springer Berlin Heidelberg 42 (2011), 181–195.

12.

Adams

, McLaren

B.M.

, Durkin

, Mayer

R.E.

, Rittle-Johnson

, lsotani

and Van Velsen

, Erroneous examples versus problem solving: Can we improve how middle school students learn decimals? In Miyakem

, Peebles

and Coppers

R.P.

(Eds.), Proceedings of the 34th Meeting of the Cognitive Science Society 12(4) (2012), 1260–1265.

13.

Kumar

M.S.

and Kumaraswamy

Y.S.

, An improved support vecto machine kerel formedical image retrieval system, International Conference on. IEEE (2012), 257–260.

14.

Tangelder

and Veltkamp

, Survey of content based 3D shape retrieval methods, Mutimedia tools and Applications 39(3) (2008), 447–471.

15.

Zhang

, lin

and Zhang

. Support vecto machine learning for image retrieval, IEEEInt’lconf On Image processing 5(20) (2001), 721–724.

16.

Stricker

M.A.

, Markus Orengosimilarity of color images Storage and Retrievalfor Image and Video Databases, International Society for Optics and Photonics 19(5) (2019), 381–397.

17.

Dzeng

R.J.

, Lin

C.T.

and Fang

Y.C.

, Using eye-tracker to compare search patterns between experienced and novice workers for site hazard identification, Safety Science 82(3) (2016), 56–67.

18.

Zhu

, Wu

, Huang

, et al., Fast open-world person re-identification, IEEE Transactions on Image Processing 27(5) (2017), 2286–2300.

19.

Koulayev

, Search for differentiated products: identification and estimation, The RAND Journal of Economics 45(3) (2014), 553–575.

20.

Howarth §²., Rüger

, Evaluation of texture features for content-based imageretrieval, Image and Video Retrieval (2004), 326–334.

21.

Tong

, Wong

W.K.

and Kwong

C.K.

, Differential evolution-based optimal Uabor filter model for fabric inspection, Neurocomputing 173 (2016), 1386–1401.

22.

Liu

, Wen

, Su

, et al., Automatic slub detection using Gabor filters, 103 International Journal of Clothing Science and Technology 20(4) (2008), 214–221.

23.

Kumar

and Pang

G.K.H.

, Defect detection in textured materials using Gabor filters, IEEE Transactions on Industry Applications 38(2) (2002), 425–440.

24.

Kumar

and Pang

G.K.H.

, Defect detection in textured materials using optimized filters, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics) 32(5) (2002), 553–570.

25.

Bodnarova

, Bennamoun

and Latham

, Optimal Gabor filters for textile flaw detection, Pattern Recognition 35(12) (2002), 2973–2991.