Academic risk prediction is a hot topic in the field of big data in education that aims to identify and help students who experience great academic difficulties. In recent years, the use of machine learning algorithms to achieve academic risk prediction has garnered more attention and development. However, most of these studies use static statistics as features for prediction, which are slightly insufficient in terms of timeliness. To be able to capture students who have difficulties in course learning in a timely manner and to improve the academic performance of school students, this paper proposes a method based on multivariate time series features to predict academic risk. The method includes three steps: first, the multivariate time series feature is extracted from the interaction records of the students' online learning platforms; second, the multivariate time series feature transformation model ROCKET is applied to convert the multivariate time series feature into a new feature; third the new feature is converted into a final prediction result. Comparative tests show that the proposed method has high effectiveness.