Keyword: Artificial Inteligence, Spatial Database, Spatial Modeling, Web Developement, Data Analysis
2022
Research
This project built a hedonic regression model to predict housing prices in Mecklenburg County, North Carolina, using over 44,000 property transactions and 40 explanatory variables. The goal was to understand spatial, socio-economic, and amenity-based drivers of housing value while testing the model’s generalizability across different neighborhoods.
Accurate housing price prediction is essential for urban planning, taxation policy, and equitable community development. Overvaluation in low-income areas can lead to higher property taxes and displacement, while undervaluation may reduce investment and services. This model aimed to create a data-driven framework to better understand such spatial inequities.
Data from the City of Charlotte Open Data Portal and the Mecklenburg Quality of Life Explorer were processed using R. Independent variables were grouped into three categories: internal housing characteristics, public service and amenity access, and spatial context. The model was trained on 80% of the dataset and validated on the remaining 20%, applying feature selection, correlation analysis, and spatial diagnostics including Moran’s I to assess clustering and spatial bias.
Developed in 2022 as part of the Master of Urban Spatial Analytics (MUSA) program at the University of Pennsylvania, this project was completed independently under the guidance of Professors Elizabeth Delmelle, Michael Fichman, and Matt Harris.
The final model achieved an R² of 0.70, explaining 70% of price variation with a Mean Absolute Error (MAE) of $115,126 and a Mean Absolute Percentage Error (MAPE) of 28%. Cross-validation confirmed stable generalizability (ΔMAE ≈ $900 across folds). Spatial analysis showed limited clustering of residuals (Moran’s I = 0.19), indicating moderate spatial bias. While not accurate enough for taxation, the model provided planners with an interpretable and spatially robust tool for understanding housing market dynamics.