Honor thesis

So senior year of undergrad, which means a season for doing honor thesis. For one thesis, I use Vim to type the paper in Latex, where the analysis is done in Python with script also in Vim, and the topic is predicting futures prices in a very short term horizon (or high frequency trading to be fancy). And the data is almost big (2 Gigabytes of tick data of E-mini S&P500 futures).  

You don't get much cooler than that. Unless you also use some statistical learning tools to maximize out of sample testing.  All of the fanciness but the model is still not making profit. So not that fancy. But it comes close. So we're getting there.

The other paper I'm working on can be again fancily called unsupervised learning,  or in a more correct term Multidimensional Scaling, or in a less scary term it is similar to Principal Component Analysis. It is also done in Python, requires a small understanding of XML, and deal with congress voting data. So quite cool too.

Now enough the fancy sounding and back to the ugly Terminal.