Assessing Models and Simulations of Epidemic Infectious Diseases

The Data Exchange Podcast: Bruno Gonçalves on reading epidemic models and time-series analysis.

Subscribe: iTunes, Android, Spotify, Stitcher, Google, and RSS.

In this episode of the Data Exchange I bring back Bruno Gonçalves, a data scientist working at the intersection of Data Science and Finance. Bruno was a guest on this podcast in April, when the COVID-19 cases were spiking in his home base in NYC. Prior to shifting over to data science, he spent several years as a researcher focused on mathematical models in Epidemiology – a field with a rich history dating as far back as the 1920s.  I wanted to bring him back to get an update on the mathematical models being used to model the global pandemic.

We discussed how the epidemiological models have performed, and adjustments modelers may have made to their models. We also discussed efforts to augment models with alternative data sources (see “An Early Warning Approach to Monitor COVID-19 Activity with Multiple Digital Traces in Near Real-Time”). We closed with a discussion of time-series modeling, and the use of machine learning and deep learning in time-series.

Here are a few of Bruno’s recent posts on CoVID-19:

• CoVID-19: Visualizing individual patient data (How to analyze 2.3 Million individual patients data)
• CoVID-19: The first truly global event (A period that will change the way we see each other)
• Visualizing the spread of CoVID-19: A How-To Guide
• Accompanying code can be found on GitHub.
• You can view a video version of this conversation on our YouTube channel.

Bruno was kind enough to assemble the following pointers to some of the references we discussed during our conversation:

• The group at LANL working on agent based models for epidemic spreading.
• Publications of Marco Ajelli’s group.
• Neil Ferguson’s team at Imperial College in London.

Subscribe to our Newsletter:
We also publish a popular newsletter where we share highlights from recent episodes, trends in AI / machine learning / data, and a collection of recommendations.

Related Content:

• Bruno Gonçalves: “Computational Models and Simulations of Epidemic Infectious Diseases”
• Amy Heineike: “Machines for unlocking the deluge of COVID-19 papers, articles, and conversations”
• Chris Nicholson: “Next-generation simulation software will incorporate deep reinforcement learning”
• Ameet Talwalkar: “Democratizing Machine Learning”
• Christopher Nguyen: “Designing machine learning models for both consumer and industrial applications”
• Chris Wiggins: “Viewing machine learning and data science applications as sociotechnical systems”

[Image: Art from Pinhole Coffee, photo by Ben Lorica]