A Big Data Based Method for Pass Rates Optimization in Mathematics University Lower Division Courses

Abstract In this paper a big data based method is presented, for the enhancement of pass rates in massive lower division courses of mathematics at University level. We propose the student-lecturer match as the cornerstone of our optimization process. First, we use the available historical data to compute the success probabilities of students-lecturer within profile segments. Next, using integer programming models, the method finds the optimal pairings of students-lecturers, in order to maximize the success (pass) chances of the students’ body. Throughout the paper, we will present in parallel the examination of our method as an economic process, as well as its importance for public universities in underdeveloped countries.