Today it is commonplace to hear of network security breaches in the news. The Stuxnet worm which disrupts the nuclear development of an entire country and the buffer over-read Heartbleed bug are only a sampling of many attacks that we know about. The traditional TCP/IP protocol was built on a set of assumptions that do not necessarily hold anymore. To this end, the goal of our research is to catalog the current literature on network-disabling attack planes and then explore novel approaches to bringing down or disabling the intended operation of a networked system, such that development of new clean slate architectures can be done with the security concerns in mind.

During parts of my freshmen and sophomore years, I worked on using novel machine learning techniques to predict diseases associated with the time evolution of autism in children. This past summer I interned at Yelp on their ad targeting team, and worked on a tool (which we open sourced) that can optimize parameters from continuous spaces to assist with A/B testing.