Educational Data Mining (EDM) involves the extraction of concepts and similar useful information from data sets that store information about academic work. EDM incorporates a toolkit, techniques, and ways of designing research that can automatically reveal correlations and patterns from substantial data sets harvested within educational environments. Making predictions of student attainment has become a significant challenge as educational data sets contain so much data. However, Learning Outcome Assessments (LOA) are crucial both for assessing and effecting improvements in teaching and learning quality and to guide individual students' development. This research aims to make student performance predictions more efficient and accurate, having the aim of offering educational institutions information crucial to improvement of learning outcomes as early as possible. This paper employs regression and several machine learning methods for the development of learning models that can offer accurate predictions of student GPA. Additionally, a number of attribute evaluator methodologies were employed for the identification of those elements that significantly influence a student's total performance.