Customized review stride frequency recommendation algorithm

The invention discloses a customized review stride frequency recommendation algorithm, and relates to the technical field of memory algorithms, and the algorithm comprises the following steps: S1, constructing a database; s11, aiming at different learning subjects, constructing different subject groups; s12, filling knowledge points of the subject group, grading the knowledge points from simple to difficult, and respectively recording the knowledge points as grades 1 to 10; s2, obtaining knowledge point memory data; s21, the subject group knowledge points are pushed to clients, and the pushing interval is adjusted according to the levels of the knowledge points; s3, constructing a regression curve; s31, the knowledge points are pushed by multiple persons, and the levels of the knowledge points are adjusted; s32, obtaining self-scoring values of multiple groups of clients, and fitting to obtain a regression equation; s4, customized pushing is carried out; s41, aiming at customer knowledge point training, adjus