I work on data-driven security. My research objective is to provide an evidence based foundation for security, by building defenses grounded in a rigorous understanding of real-world adversaries.
I conduct empirical studies of adversary behavior, I build machine learning systems for detecting malware and attacks, and I study the security of machine learning in adversarial environments. I also have a good knowledge of the security industry, having worked for 2.5 years at Symantec Research Labs. There, I built WINE, one of the first platforms for sharing field data collected by the security industry with academic researchers. In my most cited paper I measured how long zero-day attacks go on undiscovered in the wild; this measurement was made possible, for the first time, by the WINE platform.
My research has been featured in the Research Highlights of the Communications of the ACM and has been widely cited in the media, for example in The Economist, the MIT Technology Review, Forbes, and The Register. I also enjoy giving TED-style talks, to explain our work to broad audiences.
I have a Ph.D. from Carnegie Mellon University and undergraduate degrees from the Ecole Polytechnique and the “Politehnica” University, Bucharest.
My academic CV: [PDF]
At a Glance
Recent papers (selection): [USENIX Security’22] [NeurIPS’21] [USENIX Security’21] [ICML’21] [ICLR’21] [ICLR’20] [USENIX Security’19] [ICML’19] [NeurIPS’18] [CCS’18] [USENIX Security’18(a)] [USENIX Security’18(b)] [USENIX Security’18(c)] [WEIS’18] [Euro S&P’18] [CACM’18] [CCS’17] [NDSS’17] [CCS’16] [Onward!’16] [ICSME’16] [CCS’15] [USENIX Security’15] [OAKLAND’15] [IMC’14] [RAID’14] [ASONAM’13] [CCS’12]
Systems and data sets released:
- https://www.exploitability.app/: Expected Exploitability for recent vulnerabilities
- http://hardwarefail.ml/: Hardware attacks against machine learning
- http://shallowdeep.network/: A generic multi-exit architecture for deep neural networks
- ChainSmith: IOCs with killchain phases
- FeatureSmith: Automatic feature engineering for malware detection
Program Committees I am on this year: Oakland’22, and USENIX Security'22
I am chairing the program committee of RAID'22
Student(s) who will join my group in Fall'22: Shoumik Saha, Kamala Varma, and Sriram Balasubramanian
Email: firstname at umd dot edu