Welcome to the official webpages of the unofficial Predictive Analytics Group (PAG)! We work within the Associate Provost Office of Enrollment and Academic Strategic Planning at Michigan State University.
Our work is at the intersection of potential and current students’ data to ask and answer questions about potential students’ behaviors and outcomes. Using our combined expertise in economics, analytics, and predictive modelling, we provide a unique perspective on student-centric quandaries. Currently, we are studying economic mobility and gaining a deeper understanding of applicants.
We believe in creating and fostering a supportive and inclusive environment for our members to thrive and advance their career. To support our efforts, we developed and adopted an evolving collection of Guidelines & Policies and prompts about how each of us likes to be mentored.
We have informal group meetings every couple of weeks where the current and past members come together to share updates about current and past work. Everyone is welcome to attend! Please contact Emilio for day, time and location (usually a local coffee shop). Sometimes we meet on Zoom; email Emilio for the link.
Current Members
We believe that a small group with a diverse collection of viewpoints and experiences is the optimal approach to many types of research questions. In the three short years, PAG achieved extraordinary success. Building a deposit likelihood model capable of identifying applicants likely to deposit as early as the January before the deposit deadline. Additionally, they designed and implemented a modular feature pool creation application that significantly reduced the amount of time from feature concept to evaluation in a model.
Emilio is forever grateful for the opportunity to work with Seth and Steve. Their diverse experiences and skills, coupled with their desire to solve complex problems and ask challenging questions, made working with them a joy. Unfortunately, in June/2025, due to budget cuts, Seth’s and Steve’s contracts were not renewed. Emilio is devastated.
Emilio Xavier Esposito: Emilio is an Academic Specialist in the Associate Provost Office of Enrollment and Academic Strategic Planning and explores novel ways student data relates to various outcomes. Emilio’s background in early stage drug discovery, specifically computational chemistry and biochemistry. Emilio leads the development of the deposit and cancellation likelihood modelling feature pools and models, identification of new recruitment communities using informatics-based approaches, and ad hoc case studies related to predictive analytics. Emilio enjoys reading, walks with his family, and day dreaming about working in the WNBA or NWSL. [ LinkedIn | Website | Bluesky ]
Past Members
Seth Tanner Walker: Seth is a Graduate Student in the Department of Economics in the College of Social Science working towards a Ph.D. in Economics. Seth worked on the development of XGBoost models to predict the likelihood of applicants to deposit and feature engineering for those models. He also explored the impact of the Spartan Advantage program. Seth earned dual Bachelors degrees from James Madison University in Quantitative Economics & Mathematics and Music Performance. Seth is a huge Chelsea FC fan and enjoys hiking in far away lands. [ LinkedIn | Website | Social ]
Steven R Miller: Steve is an Assistant Professor in the Department of Agricultural, Food, and Resource Economics in the College of Agriculture & Natural Resources and specializes in applied economic methods for forecasting and impact analysis. Steve is also the Director of the Center for Economic Analysis. Steve explored the movement of college students within and from their home state and lead the development of community based features for the deposit likelihood models. Steve is an avid photographer and spends many nights chasing the Northern Lights and several early mornings photographing fields awakening after a night of sleep. [ LinkedIn ]