https://doi-002.org/1018/17587846047300
Xiaoyu Wen
School of Modern Technology, Xi’an Traffic Engineering Institute, 710300, Xi’an, Shaanxi, China
Email: 18729938992@163.com
Abstract: With the rapid development and application of computer technology, the research on information teaching ability of colleges and universities is one of the main research works in colleges and universities at present. The teaching of colleges and universities needs to be transformed by information technology. Because the teaching methods of the old-fashioned teaching can’t satisfy the growing demands for teaching. After training the model, it can be used to predict and evaluate new teaching ability data. Specifically, new teacher information, student information, course information, etc. can be input, and then the evaluation results of the teacher’s teaching ability can be output. At the same time, this model can also be used to analyze and compare the impact of different teaching methods, teaching content, teaching attitudes, etc. on teaching ability, in order to further improve teaching quality. In this study, neural network algorithm was adopted to evaluate teaching ability. According to the error inverse method of neural network algorithm, the algorithm model of university information teaching was constructed. The algorithm model was constructed for the knowledge needed in colleges and universities. From the research results, the research of this paper is feasible.
Keywords: neural network; university informatization; teaching ability
- Introduction
The informationization teaching ability of universities refers to the ability of teachers to use information technology methods for teaching, including teaching design, teaching resource development, teaching implementation, teaching evaluation, and other aspects. In order to comprehensively evaluate the informationization teaching ability of universities, a neural network method can be used. Specifically, Multilayer perceptron (MLP) neural network model can be used for comprehensive evaluation [1]. The model includes an input layer, a hidden layer, and an output layer, where the input layer represents the evaluation indicators and the output layer represents the evaluation results. In the training process, the Backpropagation is used to optimize the model, so that the output results are as close as possible to the actual results [2]. When determining evaluation indicators, representative indicators can be selected based on the characteristics of information technology teaching capabilities in universities, such as teaching resource quality, teaching implementation effectiveness, and teaching effectiveness evaluation [3]. When determining the evaluation results, they can be divided into several levels based on the actual situation, such as excellent, good, average, and poor. When training the model, historical data can be used as training samples, and expert evaluation results can also be used as training samples [4]. After the model training is completed, the model can be used to comprehensively evaluate the informationization teaching ability of universities. It should be noted that although neural network models have good generalization and adaptability, they also need to be adjusted and optimized according to actual situations. At the same time, when using neural network models for comprehensive evaluation, it is also necessary to consider the weight of evaluation indicators and the rationality of evaluation results [5]. Human beings are now faced with an era in which information technology is widely used. Based on the traditional teaching methods, the application of information technology in education in Colleges and universities means the use of information technology in Colleges and universities to create a completely open physical and cultural atmosphere [6]. In order to make full use of and integrate the comprehensive knowledge into the teaching of colleges and universities, to explore a new type of university teaching methods and teaching ability research, to improve teaching quality and efficiency of teaching in universities, colleges and universities also need a new teaching method to transport the talents needed for the society [7]. College students can adapt to the real environment of society. At present, colleges and universities also need to further open the teaching resources and create conditions for their students to enhance their competitiveness [8].
At present, the construction of education informationization in colleges and universities is the product of the current information age. The informationization of colleges and universities is the application of modern information technology, in-depth development, teaching information resources to the current field of colleges and universities education. Therefore, it is necessary to carry out information technology teaching in colleges and universities [9]. The construction of university teaching ability in this study is based on the neural network algorithm proposed by scientific research group in 1986 led by Rumelhart and Mccelland. The algorithm derives a multi-layer feedforward network based on the error back propagation algorithm. The neural network algorithm is studied by simulating the neural operation of humans and higher animals [10].
- Related Work
Neural network models require a large amount of data and computational resources to support, so it is necessary to accumulate sufficient experience and data in teaching practice, and make appropriate adjustments and optimizations [11]. Meanwhile, neural network models are only one type of teaching tool and cannot completely replace the role of manual teaching. It is necessary to combine interaction and collaboration with teachers and students in order to achieve the best teaching effect [12]. Neural network models can be used to analyze students’ learning habits and interests, thereby designing personalized teaching plans based on the characteristics of different students. Neural network models can be used to analyze the relationship between teaching content and teaching methods, in order to develop high-quality teaching resources suitable for different student needs [13].
After years of research in medicine and physiology, people found out how to construct nerves. After constructing a neural network, this algorithm can quickly make feedback and search for the knowledge that people need, which is a very convenient way [14]. Therefore, by collecting more teaching resources from outside, the teaching methods of colleges and universities are updated with the times, or they will be eliminated by the times [15]. With the rapid development of information technology and communication technology, most of teaching in colleges and universities is conducted through the Internet. Therefore, as early as 60 years ago, the developed countries have paid attention to the application of information technology in education, a qualified university needs to keep pace with the times and improve teaching methods in the teaching of education [16]. Neural network is a machine learning technique that mimics the human nervous system and can be used to establish various types of models, including classification, regression, clustering, etc. In the information management of colleges and universities, the neural network can be used to establish the management and evaluation system of educational information resources, so as to improve the management efficiency of educational information resources and the ability of Information assurance [17]. For example, neural network method can be used to establish the mathematical model of the educational information resource management evaluation system, take the evaluation index as the input of the neural network, and take the management benefit of educational information resources as the output. Based on the principle of least square method, gradient search technology can be used to try to minimize the mean square error between the actual output value and the expected output value of the network [18]. Through this method, a neural network model suitable for the management of educational information resources in universities can be established to improve the scientificity and rationality of information management.
In this paper, the neural network algorithm is used, neural network is one of the commonly used algorithm models in the construction of the model, it has similar characteristics of human brain, in the information education in colleges and universities, they are consistent with people’s thinking model [19]. Therefore, the algorithm used in this paper can well meet the needs of this research. In the process of the study of the algorithm, this paper draws on a large number of foreign research data, and the construction of the model can greatly promote the development of education in colleges and universities [20]. So in the educational informationization of colleges and universities, this article will integrate the current education information through the network, which will surely develop a good neural network model.
- Methodology
3.1 Comprehensive evaluation of university informatization teaching ability based on Neural Network
The neural network was first established through the simulation of the nervous system of the human brain, so it is composed of simple information processing units, also known as the neuron structure. The interconnection of these neurons has combined to a model for the information processing algorithms needed in this paper. Because the basic unit of processing is the neuronal structure, this paper must pay attention to the improvement of neuronal performance on the basis of the previous, so as to help the research of this paper. The construction of neurons is very important, so this article improves on the original algorithm. Because the algorithm evaluation model needs to be established in this paper, it is necessary to unify these sub indexes by means of subordinate functions, and this data needs to be transformed by the following functions:
Firstly, this study needs calculate the average of the index, and the calculation function is as follows:
(1)
Then, the paper calculates the variables through different types of indicators respectively. The first is to limit the benefit type economic index.
(2)
Then this article needs to revise the index of yield returns:
(3)
The in the above formula is a function of the j-th index. After the data is unified, the learning algorithm that is unique to genetic algorithm needs to be designed, so this paper needs to test the error of result of genetic algorithm. The error of this paper is calculated as follows:
(4)
In the above formula, represents the expected output value of the network, represents the actual output value of the network, in which e is the error value between the expected output and the actual output. The distribution of the three in the model is shown in Figure 1.
Fig. 1 The distribution map of the three models
Because every genetic algorithm needs to be repeatedly calculated several times and finally come to a more accurate algorithm evaluation mode. One of the main functions of neural network is the self-learning. The learning process of the error backpropagation algorithm consists of two processes: the forward propagation of information and the back propagation of error. Each neuron in the input layer is responsible for receiving the input information from the outside world and passing it to the neurons in the middle layer. The middle layer is the internal information processing layer, which is responsible for the information transformation. According to the demand of information change ability, the middle layer can be designed as single hidden layer or multi-hidden layer structure. After the information of the neurons passing through the last hidden layer to the output layer is further processed, a forward propagation process is complete. The output layer outputs information processing results to the outside world. When the actual output does not match the expected output, it enters the error propagation phase. The error is passed through the output layer, and the weight of each layer is corrected according to the error gradient descent, and is transmitted back to the hidden layer and the input layer. The process of forward and backward propagation of information and back propagation of errors are the process of continuously adjusting the weights of all layers and also the process of neural network learning and training. The process continues until the error of network output is reduced to an acceptable level or until a predetermined number of learning is reached. In this paper, the error correction can be carried out through the learning process of neural network. This paper can set the self-learning through the following mathematical expressions: So this article specially designed a new formula.
(5)
(6)
In the formula, is the network coefficient of the neural network designed in this paper, is the network input reference coefficient of the network learning rate, is the continued product of membership degree of input parameter, e is the modification factor. Through this coefficient correction algorithm, in this paper, after several operations of the algorithm, this paper modifies the checking coefficient of the algorithm through correction algorithm. In this way, a neural network algorithm that is in line with the evaluation method of education informatization in colleges and universities can be derived. The calculation data table is shown in the following Table 1.
Table 1 The statistics data of evaluation methods
| Data input value | Network learning rate | Degree of membership | Correction factor | Network coefficient |
| 100 | 50% | 44 | 45 | 69% |
| 200 | 49% | 45 | 49 | 78% |
| 300 | 48% | 37 | 38 | 65% |
| 400 | 47% | 39 | 41 | 70% |
| 500 | 55% | 42 | 45 | 75% |
From the data in the above table, the algorithm studied in this paper can be seen. In the study of the error rate of the algorithm, through the study of the above data, basic construction data of the neural network in this paper can be seen. On this basis, this article can carry out research, without the need for data collection again.
3.2 Establishment of information teaching evaluation model in Colleges and Universities
After the basic formula data processing of the algorithm has been established, this paper needs to add the comprehensive evaluation of the university information technology teaching ability based on the neural network to the algorithm model. Nowadays, the teaching of teaching informationization in colleges and universities is still in a new stage. Because our country colleges and universities are still at the embryonic stage in the research of informationization teaching, this article needs a good evaluation method for each university’s education informationization. The education informatization of university promotes the cultivation of talents, which is the foundation of the application of information technology in education. For the built information environment and information resources, this article is to use, and second, to make good use of it. Research at home and abroad shows that the main reason behind the slow pace of education informationization is neither technology nor economics, but the shackles of outdated education system and education concept. Only by changing the traditional teaching concept, curriculum view, teaching mode view, teaching organization view, education informatization has the real significance. In order to reflect the requirements of quality education, innovative qualified personnel should include the following qualities: ideological and political quality, physical and mental quality, cultural quality (including information literacy), professional quality, innovative quality. Education information construction should be directly served on this premise. The change of teaching mode should pay attention to the instructional design of informationization. The teaching mode should be transformed from the traditional mode characterized by “imparting knowledge — accepting knowledge” into an open and multi-modal modern teaching mode that is compatible with the information age and relying on the internet and multimedia. The teaching model should be reasonably combined, applied comprehensively and optimized in an overall way. It is the key to the success of the education informatization in colleges and universities to embody the modern educational ideas and concepts and pay attention to the teaching design of information. It is necessary to pay attention to the integration and complementation of advantages. Traditional campus teaching has its own unique advantages, such as the influence of teachers’ personality charm, influence of campus culture, interpersonal skills, emotional goals and the cultivation of motor skill goals. People should not blindly emphasize the application of modern information technology and neglect these things that are also excellent. After all, education is people-oriented. Therefore, this article has added these evaluation systems into the evaluation system of education informatization. The algorithmic model of this article is shown in Figure 2.
Fig. 2 Simplified flow chart of algorithm
Based on the above model, this paper adds the learning sample, which consists of two parts: the input sample and the output sample. That is, the membership function value of the subject to be evaluated. In this paper, the trained samples are input into the network model. After the model training, the neural network contains the teaching data of each school after the original sample model. This paper stores the information into the value of the tradeoff of neural networks, and then makes a systematic evaluation of the object system, so the following algorithms are obtained: The attribute value of the indicator is digitized, the membership function value of each index is obtained, and then the structure parameters of them are adjusted with the structure of the neural network.
- Result Analysis and Discussion
After designing the algorithm model, the algorithm needs to be tested, not only to test the correctness and performance of the algorithm, but also to conduct the feasibility test of the algorithm model in the process of education informatization in the actual colleges and universities. Based on the chapter model, this article prepared core diagnostic code in Java. In this paper, based on integration of information education resources in colleges and universities of neural network, the evaluation model is checked through the test of 15 universities in our country. This article uses MATLAB, and neural network running toolbox. This original data was tested for validity from the history of education in our country. Through the sub-index quantified membership function value and the expected value of this paper, the paper records the calculated value of the algorithm, as shown in Table 2 below:
Table 2 The value of membership function and budget expectations
| University code | Index | Yi1 | Yi2 | Yi4 | Yi5 | Yi6 | Yi7 | Yi8 | Ji |
| w | 0.2 | 0.15 | 0.15 | 0.1 | 0.2 | 0.15 | 0.05 | ||
| 1 | 0.481 | 0.094 | 0.275 | -0.259 | 0.169 | -0.099 | -0.099 | 0.103 | |
| 2 | -0.086 | -0.254 | 0.330 | -0.185 | -0.060 | 0.905 | -0.332 | -0007 | |
| 3 | 0.334 | 0.363 | 0.215 | 0.060 | 0.905 | 0.652 | 0.126 | 0.445 | |
| 4 | -0.637 | -0.330 | -0.064 | -0.179 | -0.786 | -0.410 | -0.213 | -0.378 | |
| 5 | 0.481 | 0.094 | 0.275 | -0.259 | 0.169 | -0.099 | -0.099 | 0.103 | |
| 6 | -0.185 | -0.060 | 0.905 | -0.179 | -0.786 | -0.410 | -0.330 | -0.064 | |
| 7 | 0.094 | 0.275 | -0.259 | 0.169 | 0.094 | 0.215 | 0.060 | 0.905 | |
| 8 | 0.263 | 0.129 | -0.318 | 0.107 | 0.228 | 0.031 | 0.006 | 0.005 | |
| 9 | 0.275 | -0.259 | 0.169 | -0.099 | -0.099 | 0.103 | -0.786 | -0.410 | |
| 10 | -0.086 | -0.254 | 0.330 | -0.185 | -0.060 | 0.905 | -0.332 | -0007 |
According to the weight value in the knowledge base of this article and the threshold set in this paper, the threshold of neural network is assigned. The output function value of the hidden node is set. Through the above construction process, this paper can establish a preliminary neuron model. On this basis, the interaction and application of neurons can be carried out. Artificial neural networks reflect several basic characteristics of human brain function, but not a vivid description of biological systems, just some kind of imitation, simplification and abstraction. Compared with the digital computer, artificial neural network is closer to the human brain in terms of composition principle, function and so on. Instead of executing the calculations step by step in the given program, it can adapt itself to the environment, summarize the rules, and perform some kind of computation, recognition, or process control. Neurons, like other types of cells, include cell membranes, cytoplasm, and nucleus. However, the shape of the nerve cells is rather special, with many protrusions, so it is divided into three parts: cell bodies, axons and dendrites. There are nuclei in cells, and the function of projections is to transmit information. The dendrites act as protuberances that introduce the input signal, and the axon is the output and has only one. So the error value calculated in this paper is shown in Figure 3.
Fig. 3 The error of expectation and actual output
It can be seen from the table that the algorithm studied in this paper has high accuracy of the previous research algorithm. Through the above calculated data, the paper adjusts the arithmetic parameters in the algorithm model and modifies the quantified values of the indicators. Finally, this paper draws the specific operation value of this algorithm. By means of training, the model in this paper is modified. After the algorithm is run, the model is modified accurately. Then, this paper needs the output value and the expected value of the neural network model. In this paper, the data error of the algorithm is tested in the following Table 3.
Table 3 Neural network model error algorithm
| University code | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
| Training data | 0.102 | -0.071 | 0.447 | -0.381 | 0.008 | 0.295 | -0.376 | -0.544 |
| Desired output | 0.103 | -0.071 | 0.445 | -0.037 | 0.008 | 0.296 | -0.377 | -0.543 |
| Relative error% | 0.97 | 0 | 0.45 | 0.79 | 0 | 0.34 | 0.27 | 0.18 |
| Simulation scheduling | 4 | 1 | 7 | 10 | 3 | 6 | 8 | 2 |
| Major sorting | 4 | 1 | 7 | 10 | 3 | 6 | 8 | 2 |
From the above table, it can be seen that the maximum error between the output value and the expected value obtained by this algorithm model is 0.88%. The ranking of the comprehensive evaluation of the information resources of colleges and universities is consistent with the expert ranking. Therefore, this comprehensive evaluation of neural network learning resource management is combined with the expert knowledge combined with this article, and makes further changes on this basis.
- Conclusions
With the rapid development and application of computer technology, the research of information technology teaching ability in colleges and universities is one of the main research works in colleges and universities at present. The teaching of colleges and universities needs to be converted into information technology. Because the old fashioned teaching methods can’t meet the increasingly increased teaching needs. In this study, the neural network algorithm was used to evaluate the education in this paper. According to the error inverse method of neural network algorithm, the algorithm model of information teaching in colleges and universities was constructed, and the knowledge needed in colleges and universities was used to construct the algorithm model. Human beings are now faced with an era in which information technology is widely used. Based on the traditional teaching methods, the application of information technology in education in colleges and universities means the use of information technology in colleges and universities to create a completely open physical and cultural atmosphere. In order to make full use of and integrate the comprehensive knowledge into the teaching of colleges and universities, to explore a new type of university teaching methods and teaching ability research, to improve teaching quality and efficiency of teaching in universities, colleges and universities also need a new teaching method to transport the talents needed for the society, so that students in colleges and universities can adapt to the real environment of society, and now colleges and universities also need to further open the teaching resources, so as to create conditions for students in schools to improve their own competitiveness.
References
[1] Ma S. Fuzzy Comprehensive Evaluation Method-based Big Data Era Chinese Sports Brands Independent Innovation Ability Research. Open Cybernetics & Systemics Journal, 2015, 9(1): 2290-2294.
[2] Li X, Zhou D. Research on comprehensive evaluation of financial innovation ability of the national-level new areas based on grey correlation. Journal of Shanghai Jiaotong University (Science), 2016, 21(2): 204-209.
[3] Dai L, Li J. Study on the quality of private university education based on analytic hierarchy process and fuzzy comprehensive evaluation method1. Journal of Intelligent & Fuzzy Systems, 2016, 31(4): 1-7.
[4] Chen J F, Hsieh H N, Do Q H. Evaluating teaching performance based on fuzzy AHP and comprehensive evaluation approach. Applied Soft Computing, 2015, 28(C): 100-108.
[5] Yue F. Research on the Comprehensive Strength Evaluation for Universities based on the Fuzzy TOPSIS Method. International Journal of Emerging Technologies in Learning, 2017, 12(8): 36.
[6] Yi Zuo. Research on Evaluation Method of Nitrogen Oxide Control Technologies Based on Grey Multi-Hierarchy Comprehensive Evaluation Method for Coal-Fired Power Plants. Advances in Environmental Protection, 2015, 05(5): 129-139.
[7] Wang X, Zhang H, Li M, et al. Research on Interference Optimization Mechanism of Wireless Routing Signal Transmission Based on Fuzzy Comprehensive Evaluation Method. International Journal of Online Engineering, 2017, 13(3): 136.
[8] Liu H, Liu K, Wang W, et al. Vulnerability Assessment of Torrent Flood Disasters Based on Variable Fuzzy Comprehensive Evaluation Method. Journal of Water Resources Research, 2017, 06(4): 332-339.
[9] Zhou T, Zhou Y. Fuzzy comprehensive evaluation of urban regeneration decision-making based on entropy weight method: Case study of yuzhong peninsula, China. Journal of Intelligent & Fuzzy Systems, 2015, 29(6): 2661-2668.
[10] Yan Z. A Building Management Evaluation Method Based on B-P Neural Networks. Open Automation & Control Systems Journal, 2015, 7(1): 1262-1267.
[11] Leung, H., Shek, D. T., Leung, E., & Shek, E. Y. Development of contextually-relevant sexuality education: Lessons from a comprehensive review of adolescent sexuality education across cultures. International journal of environmental research and public health, 16(4), 2019, 621.
[12] Barabari, P., & Moharamzadeh, K. Novel coronavirus (COVID-19) and dentistry–A comprehensive review of literature. Dentistry journal, 8(2), 2020, 53.
[13] Chen, T., Peng, L., Jing, B., Wu, C., Yang, J., & Cong, G. The impact of the COVID-19 pandemic on user experience with online education platforms in China. Sustainability, 12(18), 2020, 7329.
[14] Sturm, G., Finotello, F., Petitprez, F., Zhang, J. D., Baumbach, J., Fridman, W. H., & Aneichyk, T. Comprehensive evaluation of transcriptome-based cell-type quantification methods for immuno-oncology. Bioinformatics, 35(14), 2019, i436-i445.
[15] Osman, M. E. Global impact of COVID-19 on education systems: the emergency remote teaching at Sultan Qaboos University. Journal of Education for Teaching, 46(4), 2020, 463-471.
[16] Lim, W. G., Kim, S., Jo, C., & Lee, J. A comprehensive review of materials with catalytic effects in Li–S batteries: enhanced redox kinetics. Angewandte Chemie International Edition, 58(52), 2019, 18746-18757.
[17] Abiri, R., Borhani, S., Sellers, E. W., Jiang, Y., & Zhao, X. A comprehensive review of EEG-based brain–computer interface paradigms. Journal of neural engineering, 16(1), 2019, 011001.
[18] Sit, M., Demiray, B. Z., Xiang, Z., Ewing, G. J., Sermet, Y., & Demir, I. (2020). A comprehensive review of deep learning applications in hydrology and water resources. Water Science and Technology, 82(12), 2635-2670.
[19] Espino-Díaz, L., Fernandez-Caminero, G., Hernandez-Lloret, C. M., Gonzalez-Gonzalez, H., & Alvarez-Castillo, J. L. (2020). Analyzing the impact of COVID-19 on education professionals. Toward a paradigm shift: ICT and neuroeducation as a binomial of action. Sustainability, 12(14), 5646.
[20] Donitsa-Schmidt, S., & Ramot, R. (2020). Opportunities and challenges: teacher education in Israel in the Covid-19 pandemic. Journal of Education for Teaching, 46(4), 586-595.