Well, troops, I have good news, and I have good news.


First good news: SSL

Time to join the modern web! I'm happy to announce that angelaambroz.com is now SSL certified.


Second good news: Georgia Tech's OMSCS

I'm also happy - and nervous and excited - to announce that I was accepted into Georgia Tech's Online Master's in Computer Science. Holy moly, Morty!

This was something I had been thinking about for about four years, and actively working towards for three. My background is in economics, with an okay seasoning of math (e.g. my undergrad minor was math). But I had only ever taken one CS course previously, and it mostly talked about how to fashion a spinning hard drive out of vinyl, wood, and the blood of your clan enemies. Ho ho, kidding, little joke to convey how old I am.

Anyway, I didn't have much CS background. So, in 2015, while I was working at the Harvard Kennedy School, I started taking some CS courses through Harvard's Extension School. These courses were hugely helpful. I also started building out some additional CS foundations with free online courses, like CS50 and a bunch of other bits and bobs. But now we'll see - the proof will be in the next few years' pudding!

Big plans, Morty

I'm super pumped about the OMSCS, specifically about pivoting my career more towards artificial intelligence. I'm really excited for the courses in computer vision and AI for robotics. I've actually been actively strategizing all the courses for about 18 months (I know, too long). OMSCentral is a great resource.

Regression discontinuity okay

Joshua Goodman is a professor of public policy at HKS, and he's been doing a study on the effects of OMSCS. The tldr is: he, with co-authors Julia Melkers and Amanda Pallais, use a regression discontinuity design to create a "treatment" of admitted students (woohoo) and a "control" of those who just missed an (unknown, but it's in the paper) GPA admissions threshold. They then compare outcomes for both. The really interesting stuff is in the descriptive stats: OMSCS applicants and students are older (average age of 33), mid-career-er, male-er, American-er, and from more diverse educational backgrounds than their on-campus counterparts. Many don't apply to other graduate schools. The main finding is:

"By satisfying large, previously unmet demand for mid-career training, this single program will boost annual production of American computer science master’s degrees by about eight percent."

-- Goodman, Melkers and Pallais, 2016

EIGHT PERCENT, damn. That's a lot. That's insane.

I wasn't too surprised by the fact that OMSCS applicants are older and more mid-career and non-traditional - that seems to be how the program was pitched. It was certainly how I heard about it: in the Meetup.com circles of ambitious career transitioners in their early 30s. I'm very surprised that the OMSCS has a worse male/female ratio: only 15% of OMSCS applicants are women, compared to 25% of applicants to the on-campus program.

I'm surprised for two reasons: first, the program seems pitched towards career transitioners, and there is a big cultural push right now to get more women into tech. Ergo RailsBridge, Women Who Code, PyLadies, and so on. Maybe it's the circles I travel in, but I feel like there are a lot of women entering tech right now?

Second, the OMSCS's flexibility is perfect for, well, working parents in their 30s! In fact, one of the main reasons I was so excited by the program was that it completely fit into my personal life: I wouldn't have to uproot myself and my husband to pursue my AI/robots dream. Uprooting yourself is easier in your 20s, but once you've got the two-body problem, everything gets stickier. The OMSCS is perfect for that.

In bocca al lupo

champagne

Me, after learning I got in. Champagne!


This post is part 1 of the omscs series:

  1. Two announcements
  2. A Groundhog Day of linear algebras