The proposed project intends to use mixed methods, including machine learning techniques and in-person feedback from tenants and local partners, to predict building-level housing instability in an…

The University of Maryland, College Park was awarded a $200,000 Project Grant from the U.S. Department of Housing and Urban Development's General Research and Technology Activity program (CFDA 14.506) on March 17, 2025. The two-year study will use mixed methods, including machine learning techniques and in-person feedback, to predict building-level housing instability in an area susceptible to transit-induced neighborhood change. The ultimate goal is to understand the loss and preservation of non-subsidized affordable housing in the study area, with potential scaling to the state of Maryland. The project aims to provide data-driven insights to support housing advocacy and preservation efforts in the face of increasing property and rent prices induced by new rail transit construction.

Activities

The applicant proposes a two-year study to predict housing instability of non-subsidized buildings, a larger stock of affordable housing and less well understand than the affordable subsidized housing stock, while conducting housing advocacy outreach.

Expected outcomes

The ultimate goal of the proposed project is to use data to understand loss and preservation of non-subsidized housing in the study area and potentially includes scaling the project to maryland as a whole.1