The spread of diseases is a vital facet of healthcare research. Also of interest was the socioeconomic factor of unemployment rate and how this would affect access to healthcare and other factors that influence the spread of disease.
What it does
Creates a visualization of the percentage of all deaths that are flu deaths as well as the unemployment rate for 121 United States cities. With weekly data ranging from 1990 to 2016, we have mapped the percentage of all deaths that are flu deaths as well as unemployment rate at that time and location. Looking to the future, we have created a K-Nearest Neighbors model to predict influenza deaths using factors such as location, year, week, and most importantly, unemployment rate.
How we built it
Challenges we ran into
Accomplishments that we're proud of
We are proud of every element of our final product. The visualization of the flu data on the U.S. map with the specific coloring related to severity, the specific city data, the week slider, and overall design of the visualization are all elements that we are proud of. An accomplishment we are particularly proud of is the K-Nearest Neighbor model and it's predictions.
What we learned
Throughout this process, we have learned how to train a K-Nearest Neighbor model and make predictions based on that data. We have also learned many new elements of web design and how to deal with all of the challenges we have encountered over the past two days. This experience has also helped us develop many new collaborative and problem solving skills, as well as how to develop under pressure and with limited time.
What's next for Where in the World is Flu-men Sandiego?
An improved prediction model with more socioeconomic data would be a wonderful improvement to Where in the World is Flu-men Sandiego. With more time, we could extend the model beyond just the flu to include the spread of other diseases. We have also considered using airport data to determine if a relationship exists between transportation and the spread of disease.