Abstract:

Student attrition and low graduation rates present a significant problem to many institutions of higher education. These low student success rates have caused significant financial and academic tolls on both students and institutions. This research proposes the creation of a sequenced neural network predictive system that follows students through the entire enrollment management funnel. The ultimate aim of this research is three-fold. First, the sequenced predictive system will assist in the development of a profile of a successful student for a particular institution. Second, the system will identify “at-risk” students early in their academic career and offer institutions means for deploying student success resources in a more useful manner. Finally, the system will follow students through their given major and provide assistance in any courses that are perceived to be “gateways” for their specific major.

AUTHOR's NOTE: This research is not entirely complete. Major components, however, have been completed, detailed within this paper, and thus demonstrate the viability and importance of this research.

Keywords:

neural networks, retention, predictive modeling, enrollment management.