Community resilience and house prices: A machine learning approach

Using community resilience data at the county level in the United States obtained from the Census Bureau, we find that improvements in community resilience are associated with an increase in real estate values. Our machine learning approach indicates that community resilience plays a significant role in shaping real estate value. Furthermore, we demonstrate that the Extra Trees Regressor (ETR) method performs the best based on the root mean squared error (RMSE) standard and is effective in predicting real estate prices in a different tested sample. Finally, we conduct a grid search, exploring various parameters to further reduce RMSE and optimize our ETR method.