Analysing students' academic performance is a sub-set of EDM that deals with the extensive data collection from an ES (education system). It aims to extract data from this information meaningfully to enable the ES. Student performance is a vital factor in calculating teaching quality in colleges. The main issue of calculating if a student will fail/pass at a specific academic discipline based on a student's grades established through the semester is complex and extremely based on several situations like; course, examinations, semester, and instructions. The research article has implemented a novel classification algorithm FCNN (Featured convolutional neural network), to calculate academic student performance. The evaluation of the FCNN and S PRAR methods, which are between, were calculated and compared to indicate the student performance. The database measured the different instances and attributes, taking “The University of Jordan, Amman.” The outcome defines that the research method attained a classification accuracy of 0.95 per cent. Lastly, this analysis describes a contribution to the early-stage classification of students at high-risk of failure and DL (deep learning) techniques

EDM (educational data mining), DM (data mining), S PRAR classifier, FCNN classification method, DL (deep learning).