forecasting energy demand for texas
team Project, assessing current forecasting for texas power grid, and building a reactive hour ahead model
demystifying camping on us public land
Developed a novel data engineering method to distill training imagery and prepare for use in a regional convolutional neural network, currently tuning model!
classification of political posts
Scraped 200,000 posts from conservative and liberal subreddits. Used natural language processing to analyze key words and bigrams for each, tuned a model to predict new posts origin with an accuracy of 87%
climate vulnerability assessment
developed R code to visualize census tract groupings determined to be most vulnerable to climate change within the context of individual measures
superfund and toxic release inventory map
sourced and aggregated data related to environmental hazards and developed an interactive map indicating relative risk of current toxic releases and historical superfund sites,
health equity maps
crafted a set of custom basemaps for future GIS work and produced a repository of indicator maps for the health equity topic page
HIV PrEP survey analysis
developed SAS code to clean, re-group, and enumerate trends resulting from a bare survey data-set identified trends in frequency tables via paper, glue, and whiteboard merkers and conducted a formal regression on medication preferences
ecological niche modeling of west nile virus transmission potential
worked with the kent county health department and developed a spatial predictive model for areas prone to increased transmission risk of west nile virus, utilizing historical infection and ecological parameters specific to the local species
