The article deals with the methods of intelligent ranking of knowledge control results in the context of digital education in mathematics. The article considers approaches to learning using hybrid intelligent learning environment and features of digital transformation of mathematics education. The methods of intellectual ranking of the results of intermediate knowledge control with regard to adaptation to the digital format of education are proposed. Machine learning algorithms are proposed for modelling the educational process, in particular, generalised clustering algorithms based on the k-means method and cascade learning of autoencoder. The developed algorithmic and software is aimed at intelligent evaluation of the results of intermediate knowledge control and identification of research potential. The conducted computational experiments allowed analysing and interpreting the obtained results. The proposed models and algorithms can be used to improve the methods of knowledge control and assessment, as well as to predict the indicators of the pedagogical process.