STOR Colloquium: Natalie Baerlocher & Shemra Rizzo (The University of North Carolina at Chapel Hill)

Likelihood of undergraduate admission to UNC based on pre-application data

Every year UNC spends a significant amount of funds on events and activities meant to attract students. The university would like to encourage the best students to apply for admission. The aim of this project is to develop a quantitative means to identify the worthy students. We forecast the likelihood of admission of a student based on the information that the university collects while in high school using two different methods: logistic regression and neural networks. This pre-application data comes from surveys at events or on the web. The best students will be the ones more likely to be admitted. We are particularly interested in exploring the admissions patterns of students from North Carolina, which represent the 80% of admitted students, and those out of state who account for 20% of the new student body and whose application process is more competitive.

Refreshments will be served at 3:30pm in the 3rd floor lobby of Hanes Hall