The Hope Algorithm: Competing to Predict Dengue Fever in the Tropics

This is a full length talk that will expand on Carter's 6 minute rapid fire talk at the Dec 19th winter party.

Talk Description:
Can you predict local epidemics of dengue fever? I am excited to share a data science competition vignette. Competitions are a great opportunity to choose your adventure. You get to enhance your data modeling skills and learn from people working on the same problem.

I'll present a snapshot of using machine learning to predict spread of dengue fever in the tropics. The competition provided 15 years of weather and disease time series data from NOAA and CDC. My solution landed a top 5% result that could be easily improved. Takeaways will focus on the unique opportunity of competitions and where to find them.

About the Speaker:
Carter Stowell is a data scientist with an M.S. in Predictive Analytics from Northwestern University. He provides data strategy and analysis at Provoc, a creative agency and B-corp based in Washington D.C. that serves visionary leaders on projects with high social impact. Carter has spent 20 years immersed in empathetic web/UX practice starting at Wired and Hotbot in the 90's. He also co-founded Oysterater.com, a top site for oyster-lovers.

Image courtesy wikipedia.org