The human cost of coronavirus has continued to mount, with more than 40.3m cases confirmed globally and more than 1.11m people known to have died. Building an intuitive dashboard visualization tracking the most updated status of Covid is beneficial to the public at large.
What it does
CovidML is a web application that visualizes the trend of COVID-19 worldwide with future spread prediction. It allows users to effectively visualize current and upcoming situations of the pandemic spread.
How We built it
Our web app is based on the Flask python framework. We obtained the dataset from sources including Johns Hopkins CSSE and New York Times. The visualizations on the plots page display the status quo of infection cases and fatalities across the world using Datawrapper tool. We produced geo-referenced plots using a combination of libraries, namely numpy, matplotlib, and cartopy. We used pandas to generate a data-frame for the dataset, and then tried different models to forecast the infection trends.
Challenges We ran into
Mining and analyzing the data to build the data visualizations was a hard step. I also spend lots of time embedding the interactive plots with the web app application.
What We learned
I learned a great amount of web development and data mining, as well as how to use datawrapper to create great visualizations. I also expanded our machine learning skills through time series forecast.
What's next for CovidML
We hope to expand our web application to cover countries all around the world, add a page that visualized the multiplex impacts of Covid, and try more modeling techniques to forecast the future trend of the pandemic.