Metacognitive training system (MTS) which is developed for Algorithm and Data Structure course facilitates students to choose, use, and evaluate different learning strategies. Therefore, students are encouraged to be self-regulated. However, the system is also expected to know the best learning strategies of students to generate recommendation that will help students to better understand their self. The recommendation is generated by using various parameters such as post-test scores, learning strategy evaluation values, learning strategy access time, number of clicks and number of summary words to determine the best level of metacognition and learning strategies for students. The learning strategy recommendation system developed using Simple Additive Fuzzy Weighting Algorithm. This method adds the weight of each criterion in each learning strategy. The system's functionality is tested using Black Box method as well as validated by experts. A user acceptance test (UAT) is also carried out to prove that the system is accepted by users and requirements have been met. The UAT resulted in an average of 82.5% which is considered as very good.