Learning Level Prediction System Based on Student Academic Ability By Looking at the Highest Similarity Value

There is a diversity of academic abilities of students in each Indonesian educational institution, each student has different learning abilities according to the level of learning. Educators in providing learning level predictions are done manually, so it takes a long time in predicting student learning levels. In this study, it was able to predict students' academic abilities in an easy and fast way. The method used in predicting learning levels is Case Based Reasoning.  This method is able to predict the student's learning level to be (1) Very Bad, (2) Bad, (3) Medium, (4) Good, and (5) Very Good. This level of learning will be used as a benchmark for educators to provide appropriate values to students.  The results of this study for the academic performance of the very poor category are 0 students, the bad category is 3 students, the medium category is 79 students, the good category is 16 students and the excellent category is 12 students. The accuracy of academic performance recommendations using confusion matrix is 91.82%.