Ongoing projects and interests
I guess stay posted as to where exactly this goes over the next while…I’m talking to some groups about what to do next in Thomas’s life (tm).
Evolution is what my Ph.D. training was in, so it’s no surprise that this is a deep well that I keep returning to. Over time my interested have broadened as I have become more skilled at data analysis and statistics in general, but at heart I am fascinated by questions that have some evolutionary hook to them.
I’m currently in Atlanta (~September-October) to work on a small, quick (famous last words, it always starts that way) project looking at machine learning applications towards schizophernic and healthy samples and seeing if we can tease apart an age-related progression.
It has some interesting aspects as it is the inverse of a “classic” machine learning problem, where you have ideally you have many samples and few features; in this case the “machine” can learn past the garbage noise. However, in this case we have the inverse we have relatively few, precious samples but zillions of features relative to them.
Data science, Big data, Twitter networks, Zika
As I was winding down my first postdoc I seriously considered life in private industry, and still do to a certain extent. To that end, I’ve read several books on machine learning and taken to learning some Spark for Big Data (tm) challenges.
After some soul searching I’ve realized that for better or worse academic research in some form is probably a better fit for me, so now I’m trying to figure out how to to put the tools I’ve picked up in the last few months (which I genuinely like) to good use. One of the realms where I’ve had a lot of fun is exploring interaction networks in Twitter via mention and retweet graphs.
As a specific application I’ve been working with Loni Hagan on the interactions and dynamics of the #Zika hashtag on Twitter and how effective health communication is in this network.