Item — HN Reader

Don't get a grad degree just yet. Try a Data Engineering job (or also a software engineering job at a data-centric startup). You'd use a lot of algorithms, machine learning, and math to do all kinds of fun stuff with data. The new position is called 'Data Scientist' these days - I would use LinkedIn / Indeed.com to look for jobs with the title 'Data Scientist'

Here are some off the top of my head that are looking for mathy programmers:

Startups: Infochimps Lookout Mobile Factual Zillabyte Chart.io KhanAcademy

Corps: LinkedIn Google etc, you should be able to at least land interviews at the big tech corps, but I"ll guarantee you'll have a lot more fun at startups

I think you should get some industry experience before going off into a pure math PhD / grad degree. You could also consider doing a Master's in Statistics or Economics, both are which are super appealing these days - not sure how much you like applied math over pure math though!

1 comments

I agreed with this post up to the point where you recommend a Masters in either statistics or economics. As a former econ student/researcher, I can say that you should not get a MS in economics or statistics if you want to do scientific computing!

1. Both are too domain-specific, and they usually lack the technical (coding) rigor you want if you're dropping lots of $$ on a grad degree. There are exceptions -- the statsmodels.py team, for example, includes an econ guy from American University -- but they're exceptions, not rules.

2. At best, either degree will get you experience in MATLAB/SAS/Stata, which are used for data analysis. This is not scientific computing, where you'll have to convert slow models to C/C++/Java etc.

Instead, if you're set on coding I would suggest a Masters in applied math or a Masters in CS with a focus on topics related to stats/ML and large-scale analysis (Bayesian inference and graphical models, sure, but don't ignore classes in numerical methods or distributed computing).