Project Breakdown

The goal of the Machine Learning Model was to ingest the inputs of all counties FIPS along with the maximum, minimum, and mean values for both air pollutants (PM2.5) and Ozone (ppb) and the cancer incidence trends associated with years 2000-2014. Separate datasets with several million rows each were aggregated and combined to form one cohesive model dataset. The model was trained to predict cancer incidence trends with the combined dataset, then it was applied to a different 14-year time slice from 2003 to 2016 predicting cancer incidence rates for counties where cancer incidence data was non-existent. Future versions of this model and web application could include the prediction of other illnesses associated with environmental pollutants.