BsidesDFW 2013 Brian Wylie

uh in this case my presentation was around uh data hacking from from a scientist perspective you know what do it mean to to look at a new uh kind of large scale security data set and apply various statistics algorithms machine learning uh to that data set and so the the presentation was really kind of a pragmatic a Hands-On approach to to using uh data analysis and statistics on security data and you know a lot of presentations talk about uh these things kind of in a theoretical aspect here you know bsides is more oriented towards practical Hands-On how do I actually do it and we provided uh you know not only Python scripts and and python modules

but you know we have a nice get repository that where people can go and download the code and try it themselves that even if you did some some gigantic parameter study where you figured out the optimal K based off of intracluster distances or something um you you really don't think about the world that way you want you know in my million uh rows of CIS log you kind of you want to sense of what are the broad categories and then you want to drill down um and you know if and and if you're flipping it around if you're looking at an individual row you want to say what's really closely related to this the the main takeaway is is that

you know as you're doing your work you want to be formal more formal around what you did how you did it and how it can be reproduced so you know that that's kind of a mantra in in a lot of the scientific communities you know how did you know what did you do how did you do it and how can someone reproduce it and and there are some really nice tools and techniques uh out there now that that help you in that

process