Stats/ML consulting (San Francisco or remote only)
May 27, 2013 3:16 PM

I'm currently the principal statistician and data scientist for a research unit at a medical school here. I've also done some consulting in the past for large companies in the medical devices and insurance fields, as well as for researchers at other institutions. I'd like to branch out a little bit from the kind of work I do here and talk to people about their data and the kinds of problems they're trying to solve with it -- that's my favorite part of the process!

My specialties involve prediction models and machine learning, particularly natural language processing, as well as data visualization. My background is in applied math and statistics.

Here's a sampling of the stuff I've worked on recently:
-- developed methods for fraud analytics of medical claims data in order to detect patterns of overuse of unnecessary and potentially harmful procedures
-- developed and validated novel mortality prediction models based on underlying topic models/natural language processing (e.g. LDA/LSI)
-- created new methods for risk stratification of patients based on electronic health record data, which allowed us to generate more intuitive visualizations of specific cohorts

In addition, I've provided ongoing methodological input for principal investigators on their projects and helped prepare manuscripts, as well as troubleshoot or refactor code as needed. I've given a bunch of talks and workshops, so if you want me to teach you something or tutor, that's also a possibility. I love teaching: it forces me to think really hard about what I know and how well I know it.. Most of my work has been done in R and Python, along with a bit of Matlab/Octave and C. I know a bit of SAS and Stata as well.

Drop me a line, tell me a bit about your data and your problem and how you think I could help out, and I'll let you know if it'd be a good fit for me.

posted by un petit cadeau to Science/Biotech