A Classification Method of Learning Dynamic Models Based On K-NN Algorithm

Learning style is an important factor affecting students’ learning style and adapting to different teaching methods. The traditional classification method is easy to be disturbed by subjective factors, resulting in insufficient accuracy of classification results. With the rapid development of artificial intelligence and machine learning technology, especially the application of K-nearest neighbor algorithm, it provides an efficient method for the automatic classification of students’ learning styles. In this study, 500 students from a school were selected as research objects. Through questionnaire survey and collection of learning behavior data, 20 characteristic variables based on VARK model, Kolb learning style theory and learning habits were extracted, and K-NN algorithm was adopted to build a classification model. The experimental results of this study show that the classification accuracy of this model is up to 85.33%, especially in the classification of visual and auditory learning styles. However, for the classification of reading, writing and kinesthetic learning styles, the accuracy and recall rate of the model are relatively low, showing certain misclassification problems.