Study on the performance evaluation of online teaching using the quantile regression analysis and artificial neural network

This paper is totally different than the research design and research method in the related literature for investigating how information technology-based reading and learning process of network distance teaching affects the assessment result in the past, that is, innovative research architecture and process is adopted. Here, quantile regression analysis is applied to the investigation of how the time and frequency of log-in curriculum, browsing teaching material, and curriculum discussion in learning process record affects the final-term assessment result of multimedia design digital teaching material subject. In depth research is done under such research architecture, and it is hoped that how each independent variable affects the final-term assessment result under different quantile can be investigated. In addition, this paper has applied new artificial neural network technology to set up expert system for teacher's assessment result in distance teaching so as to reduce teacher's teaching pressure, moreover, the result can be used as reference by general researchers and scientific education researchers. The research result shows that the use of quantile regression to analyze the influence of different variable on the teacher's final-term assessment result of distance teaching is a feasible way; FOAGRNN model, as compared to other five models, has better forecasting capability.