Reassessing Home Values in Philadelphia

Since the 1990s, it was believed that Philadelphia (and many other cities) has been spatially recentralizing itself, a process that might change, inter alia, the regional price gradient. Government and other interested parties may be interested in reassessing home values in the City. This paper details a solution I have developed that can be used to easily assess home values for either tax or insurance purposes at Philadelphia and beyond. This is achieved through the development of a standalone Python-based tool that produces two estimates: hedonic and replacement cost.

As a demonstration, ensemble learning methods were used on a filtered 2019 sample of 19,106 housing unit data from the Office of Property Assessment. Outputs include predicted output database features and output feature layers, which include the attributes and estimates for each home. Visualizations are produced in auto-generated reports in PDF format for the ease of use.

Learn more about this project.