An Intelligent Classification Method for Online Teaching Student Performance Data Based on Improved Decision Trees

At present, the extraction of intelligent classification features of online teaching student achievement data is mostly one-way structure, and the efficiency of classification is low, leading to the smaller F1 value of intelligent classification. This paper proposes the design and verification of the intelligent classification method of online teaching student achievement data based on the improved decision tree. According to the current demand for intelligent classification, the online teaching student achievement data collection is advanced, and the multi-level method is adopted to improve the efficiency of classification and extract the characteristics of multi-level student achievement classification. On the basis of this, an intelligent classification model of online teaching students' performance data is constructed by improving the decision tree, and the intelligent classification of performance data is realized with the help of collaborative filtering. The final test results show that: for the selected five intelligent classification cycles of achievement data, and for the achievement data of junior high school mathematics, Chinese and English, the final intelligent classification F1 values can reach more than 0.85, all close to 1, which indicates that the intelligent classification method designed for online teaching students' achievement data is more efficient, complete and targeted with the help of improved decision trees. It has practical application value.