Projects

Using information from the Open GIS Data portals from Atlanta and surrounding counties, guides the user through hypothetical upzoning/rezoning scenarios to illuminate the impacts of small changes in zoning code on urban sprawl and its associated costs and challenges.

After collecting and normalizing the zoning data of the majority of Fulton, Dekalb, and Gwinnett counties, calculate the areas taken up by each of several zoning classifications, such as single-family residential, low-density multifamily residential, multifamily residential, etc.

Then, present the user with various representative photographs of various housing modes, like single-family homes, townhomes, duplexes, low-rise apartments, etc. After selecting which ones they would like to see in their neighborhood, the dashboard will update to display the new extent of the three counties, as well as estimates of cost reductions in sprawl-associated costs like road maintenance, utilities, congestion, etc.

Using sonic characteristics scraped from the spotify API such as Acousticness, Danceability, Energy, etc. to predict its Spotify popularity score.

The model that was settled on was a random forest model, trained on the normalized and dimension reduced features of the most popular songs of 2023, as well as the 2010-2017 period.

A total of 13 features were analyzed, finding low correlation between the features and low ability to predict a song's popularity, at least with this specific tree model.

More extensive analysis can be found on the Github.