Summary
In this chapter, we walked you through the construction of spatial regression models to better understand the drivers of nightly Airbnb prices in NYC. We started the chapter off with a refresher on OLS regression models. Using this model, we looked at the distribution of the model’s residuals to better understand some latent spatial structures that needed to be accounted for.
In the second section, you learned how to incorporate spatially engineered proximity features into the model, which dramatically improved the model’s performance. We then introduced you to spatial fixed effects and how to use the spreg library's OLS_Regimes function to build a spatial fixed effects model, which further improved performance. Within this section, we also introduced the Chow test to ensure that the neighborhoods yielded statistically different results.
In the second section, you learned about GWR and MGWR, which are models that fit local regressions for each observation...
