Homepage: Varun Kanade

Varun Kanade
Graduate Student
CS Harvard University
Office: Maxwell Dworkin 138
Email: vkanade@fas ....

(photo by Karthik Shekhar)

I have been at Harvard University since Fall 2009 and am extremely fortunate to have Leslie Valiant as my adviser. Before joining Harvard, I was a graduate student at Georgia Tech, where I was working with Adam Kalai (now at Microsoft Research) and happily still get to work with him.

Research Interests

I am mainly interested in computational learning theory, with more focus on theoretical aspects. More broadly, I am interested in theoretical computer science. Recently, I have also been interested in evolutionary biology. I got my bachelor's degree from Indian Institute of Technology, Bombay in Mumbai, India.

Publications:

Learning Hurdles for Sleeping Experts (with Thomas Steinke) (to appear in Innovations in Theoretical Computer Science, 2012) [draft pdf]
Efficient learning of generalized linear and single index models with high dimensional isotonic regression (with Sham Kakade, Adam Tauman Kalai and Ohad Shamir) (to appear in NIPS 2011). [on arxiv] (final version will be up here soon)
Evolution with Recombination. (to appear in FOCS 2011) Conference version: [pdf]. (Full version will be up here soon.)
Evolution with Drifting Targets [pdf](with Leslie Valiant and Jenn Wortman Vaughan) COLT 2010.
Potential-based Agnostic Boosting [pdf]. (with Adam Tauman Kalai) NIPS 2009.
Reliable Agnostic Learning [pdf] (with Adam Tauman Kalai and Yishay Mansour) COLT 2009.
Sleeping Experts and Bandits with Stochastic Action Availability and Adversarial Rewards [pdf] (with Brendan McMahan and Brent Bryan) AISTATS 2009.
Life (and Routing) on a Wireless Manifold. [pdf] (with Santosh Vempala) HOTNETS 2007.
SWARM: A Parallel Programming Framework for Multi-core Processors [pdf] (with David A. Bader and Kamesh Madduri) Workshop on MTAA 2007.
Approximate Symbolic Reachability of Networks of Transition Systems. (with Sudeep Juvekar, Ankur Taly and Supratik Chakraborty) Book chapter in Next Generation Design and Verification Methodologies for Distributed Embedded Control Systems Springer. (Invited Paper)

Manuscripts:

Computational Bottlenecks for Evolvability (2011). [draft pdf]

Teaching

Fall 2011

CS 220r: Cryptography - Trust and Adversity

Instructor: Prof. M. O. Rabin

Office Hours: TBA

Spring 2011

CS 228: Computational Learning Theory

Instructor: Prof. L. G. Valiant

Office Hours: Wednesdays 6-7 pm, or by appointment.