Course selection is a crucial task which may affect greatly on student performance. Because of poor performances, numerous students have been receiving formal warnings and expulsions from universities. Clearly, a good strategy for study progress which can come from course recommendation methods really holds an important role to obtain a good study performance. In addition, early warnings that release on challeng- ing courses enable students to prepare better for such courses. The cur- rent course recommendation systems are usually conducted from marks prediction and factor analysis on marks of courses based on advancements of machine learning approaches. In this study, we propose a course recommender system by using deep learning techniques with MultiLayer Perceptron and pre-processing methods. The prediction tasks are per- formed on approximately four million of mark records at Can Tho University, Vietnam to provide recommendations on course selection to stu- dents. The proposed method reveals promising results and is expected to apply in practical cases.