We had a very different sort of research paper in journal club three weeks ago, and I was too busy to jot it down until now. Dewachi et al, Changing therapeutic geographies of the Iraqi and Syrian wars. This is certainly not our usual metrics-heavy approach, so it was good exercise to try to understand it.
Last week we had a talk from Rodrigo Moreno-Serra on Universal health coverage, equity, and health outcomes. This research used instrumental variables to show that universal health coverage is good for health. One day I will understand instrumental variables—I think there should be a simple way to explain it to combinatorialists who know some epidemiology.
It has been a while since I visited my pymc-examples repository, but I got a request there a few weeks ago about the feasibility of upgrading the Seeds Example of a random effects logistic regression model for PyMC3. It turns out that this was not very time consuming, which must mean I’m starting to understand the changes between PyMC2 and PyMC3.
Last week for IHME seminar, we heard from Kate Starbird about Crowds, crisis, and convergence: crowdsourcing in the context of disasters. It reminded me of the visual displays of quantitative information I hacked on after the 2010 earthquake in Haiti.
Did I ever tell you how the US State Department called to ask if they could use that graphic in a presentation? I thought it was a prank.
As often is the case, a recording of the talk is available online.
My grad school colleague Jure Leskovec was at UW recently, talking about a cool new model for clusters in random graphs. It was part generative model, and part statistical model, and made me miss my grad school days when I spent long hours proving theorems about such things. This statistical part of his model has a very forced acronym: BigCLAM, and also a software implementation: https://github.com/snap-stanford/snap/tree/master/examples/bigclam
Maybe one day I’ll get to do some applied network research for global health, and have a chance to use this.
IHME Seminar: Unifying the Counterfactual and Graphical Approaches to Causality via Single World Intervention Graphs (SWIGs)
Thomas Richardson gave a recent seminar at IHME about how the potential outcomes crowd can make sense of graphical models and vice versa. It also has a CSSS working paper to complement it, a trend in our recent seminars: http://www.csss.washington.edu/Papers/wp128.pdf