A Teaching Evaluation Model Based on SVM Multi Classification

Abstract

To ensure the quality of online flipped classroom teaching, we try to apply machine learning to the teaching evaluation of universities to find valuable information hidden behind the teaching evaluation data. This paper mainly discusses the SVM multi classification algorithm and establish a multi indicator system for teaching quality evaluation from five aspects. Then, with the idea of the principle of first separating the easiest classes and the idea of relative distance, the binary tree is used to train and improve SVM. The empirical analysis takes the actual teaching situation of a university as an example, provides the implementation process of the algorithm on Spark programming model, and uses the improved method to determine the weight of each evaluation index, thus ensuring the scientific nature of the teaching evaluation.