Modeling municipal or urban decisions is challenging due to the abundance of variables that guide end results. One such challenging issue is the existence of vacant lots in a city, which causes poorer standard of living for the community. As a result, reclaiming these properties and putting them into productive use is a primary concern. However, each time community leaders had to ``reinvent the wheel'' and make decisions from scratch. To this end, we propose the creation of a vacant lot model and utilizing it to provide recommendations for vacant lot conversions, providing a starting point for such decision making. We define a vacant lot model in terms of determinants to a vacant lot's impact, and evaluate the proposed method on real-world vacant lot datasets from the cities of Philadelphia, Pennsylvania and Baltimore, Maryland. Our results indicate that our prediction model performs accurately on cities with a centralized approach to vacant lot conversion.
Software Engineering (MS)
Department, Program, or Center
Software Engineering (GCCIS)
Chowdhury, Md Towhidul Absar, "A Machine Learning Approach on Providing Recommendations for the Vacant Lot Problem" (2017). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus