Cognitive diagnosis method and device for predicting student score based on breadth learning, and medium

The invention discloses a cognitive diagnosis method and device for predicting student scores based on width learning and a medium, and relates to the technical field of data mining, and the method comprises the steps: S1, constructing a student vector, a test question vector, a mastery degree matrix, a difficulty degree matrix, a distinction degree matrix and a test question knowledge point incidence matrix; s2, generating knowledge proficiency mastering degree, knowledge point difficulty degree, test question distinguishing degree and association between test questions and knowledge points; s3, optimizing the test question distinction degree through the attention residual network; s4, performing interaction among the modules to simulate a real interaction process; s5, training the non-negative full-connection neural network, and updating the mastery degree matrix, the difficulty degree matrix and the discrimination degree matrix; s6, repeatedly executing the steps S2 to S4, and updating an interaction resul