« Startups Should Make a Dent in the Universe | Main | How Facebook, Twitter, Google+ and LinkedIn will change Finance »

March 03, 2012


Feed You can follow this conversation by subscribing to the comment feed for this post.

Research paper help

I am absolutely amazed at how terrific the stuff is on this site. I have saved this webpage and I truly intend on visiting the site in the upcoming days. Keep up the excellent work



As a PhD student, I find the use of statistical models in something as novel, yet practical, as your application to be heartening. However, given my inner skeptic, well-honed after a few years in academia, I have to ask-

How much more predictive power can social network activity carry compared to actual spending habits and debt numbers? Given, as one accrues default rate information, one could build hierarchical models (homophily) based on social network activity to improve prediction above and beyond credit scores. However, isn't the lag time between having enough data on default probabilities a big barrier to entry (when compared with FICO)?

Thanks and good luck!


The Lenddo Data Science job posting are at http://lenddo.jobscore.com

Alan Samuels

Very interesting. And very true.

That said, Dilbert called it first (1996). My favorite cartoon on the power of math:

The comments to this entry are closed.