|
|
|
|
|
BEN RUBINSTEINCONTACT
|
|
RESEARCH INTERESTS
I am interested in Machine Learning and Statistics in general, and in particular:
- Security of Adaptive Systems
- Security analysis of learning algorithms in deployed systems (e.g. anti-spam, network anomaly detection, worm signature generation)
- Theory of inherent trade-offs between security and generalization
- Statistical Learning Theory/Empirical Process Theory
- Combinatorial properties of concept classes, VC dimension, sample compressibility, hyperplane arrangements
I was born in Melbourne (Australia) in 1980, received the BSc and BE(hons) degrees with majors in pure mathematics and software engineering from the University of Melbourne in 2002, and then studied towards the MCompSci in the Department of CSSE at the same university under the supervision of Rao Kotagiri and Marimuthu Palaniswami. While in Melbourne I also worked on low-level microarray analysis with Terry Speed at the Walter & Eliza Hall Institute of Medical Research. In Fall 2004 I became a first year Berkeley EECS PhD student; my advisor is Peter Bartlett. My current research interests focus around machine learning and statistics, with an emphasis on learning theory and its application to computer security; for this work I collaborate with Anthony Joseph's SecML group in the RAD Lab, and Nina Taft and Ling Huang at Intel Research Berkeley. Last summer I interned at Google Research with Phil Long and this summer I am interning at Yahoo! Research with Martin Zinkevich and Jayavel Shanmugasundaram. I was previously President of Berkeley's Computer Science Graduate Student Association.
My wife, Juliet Rubinstein, is also an Australian PhD student in Berkeley EECS.
(Erdös no. 3: Paul Erdös -> Nick Wormald -> J. Hyam Rubinstein -> Ben)
| Updated: | Thu Jun 19 10:42:36 PDT 2008 |