Application of Iterative CART Decision Tree Algorithm in Studying Influence of Early Education Curriculum on Children’s Attention Improvement

Abstract

Early education curriculum plays an important role in improving children's attention. Different early education courses have different effects on improving children’s attention. We conducted a teaching experiment in an early childhood education institution in Beijing, China, asking children to choose different early childhood education curriculum packages and carry out learning. After the experiment, the children’s attention improvement level before and after learning the curriculum package was measured. The label of the significance of the attention improvement was given as the dependent variable of the study. Secondly, several features were extracted from the curriculum package as independent variables. Then, an iterative CART decision tree algorithm was proposed to construct a dichotomous model of significant attention improvement. The classification model was tested by 10-fold cross test. The test results showed that the accuracy was 84.5%, the recall rate was 90.0%, and the F1-score was 87.2%. The accuracy of the model was acceptable.