Evaluation of Mathematical Models for Teaching Precision in the Framework of Big Data

Abstract

With the continuous development and application of data computing in the whole society in the construction of digital campus and intelligent campus of each higher education institution. In the environment of education, universities use these data well, which not only affects the orderly operation of higher education but also will become an inexhaustible power to help higher education promote the reform and innovation of education and teaching system. In this paper, we focus on the teaching operation and students' independent learning by taking students' evaluation data and students' online learning data of Y school as the research objects. We conducted a preliminary analysis and transformation of students' evaluation data of a university, eliminated the abnormal evaluation data by using the improved cosine phase dissimilarity algorithm, standardized the evaluation data by using the normalization method, and used the traditional K-modes algorithm. Based on these three problems, the traditional K-modes algorithm was improved in three aspects, including the determination of the number of clustering families, the determination of the measurement of clustering distances, and the experimental results showed that the improved algorithm was more reasonable and effective.