I’m Mike Scarpati, and I’m currently a Ph.D. Candidate at the Pardee RAND Graduate School. I’m hoping to use this blog as a way to keep myself in the habit of writing as I finish up my Ph.D., as well as hopefully share some useful knowledge occasionally.
If you would like to get in touch with me, please send me an email: mike [dot] 87 [at] gmail [dot] com.
To provide a little more background, I was born in Michigan and moved around a lot as a kid. The highlight of those moves was almost definitely Singapore, where I spent 7th through most of 11th grade. After that, I moved to Ohio, then went to school at Washington University. That was a great experience, and I was able to be on the school’s swim team, take some classes, get a job, and meet my lovely wife.
That job was at Boeing, and I transitioned from being an intern (part time during the school year) to a member of the Business Career Foundations Program. I was working in various business functions, but some of my biggest contributions were related to programming. My parents made fun of me as a kid by saying that I would spend hours finding a shortcut for a 10 minute job. That has turned out to be one of my most valuable traits! It has also been a big factor that motivated me to learn to program, and got me into machine learning.
Unlike a lot of people that go to grad school, I didn’t really make a conscious decision to leave the corporate world because I was unhappy there. I actually liked it a lot, and was on a good path. PRGS was the only school I applied to, because it built on the classes I liked in undergrad while enabling me to get exposure to a lot of different domains.
Like at Boeing, I was drawn to quantitative methods and data, and that has enabled me to work in these different domains. Project work started by building a discrete event simulation model for Army equipping, then branched into
- Learning to use EC2, EMR, and Hive (over a weekend!) in order to summarize billions of outputs for the Colorado River Basin Study
- Transitioning an internal text analysis application from research code to something more useful, stable and scalable
- Developing algorithms and back-end code for an internal document classification tool
- Modeling transportation costs faced by the Army
- Using GIS data to predict the behavior of commercial tuna fishermen in the Eastern Pacific Ocean
- Developing a Solr front-end for sensitive document collections
- Screening articles for relevance in systematic reviews
How often do you get to work on tuna fishing, document classification, information retrieval, military logistics, and medical research in the same place, at the same time?
Outside of work/school, I recently got involved with AppCity, which teaches young people to develop mobile applications. Working with Thomas, I designed the curriculum and taught the first class, which we had over the summer. It was impressive seeing what everyone was able to accomplish in such a short time!
As far as less-mentally-taxing things go, I really enjoy hanging out with and training my dogs and playing video games. I’ve also recently started to make wine, and will post an update on that once it gets far enough along.