Home

Research

Classes

Cool Stuff

Research Interests

Machine Learning, high dimensional statistics. I am currently focusing on manifold learning techniques.

Publications:

Nakul Verma, Samory Kpotufe, Sanjoy Dasgupta. Which spatial partition trees are adaptive to intrinsic dimension? Twenty-Fifth Conference on Uncertainty in Artificial Intelligence (UAI), 2009. [pdf] [poster]

Yoav Freund, Sanjoy Dasgupta, Mayank Kabra, Nakul Verma. Learning the structure of manifolds using random projections. Twenty-First Conference on Neural Information Processing Systems (NIPS), 2007. [pdf] [poster] [software]

Sanjoy Dasgupta, Daniel Hsu, and Nakul Verma. A Concentration Theorem for Projections. Twenty-Second Conference on Uncertainty in Artificial Intelligence (UAI), 2006. [pdf] [poster]

Research Exam: Nakul Verma. Mathematical Advances in Manifold Learning. Survey, 2008. [pdf] [slides]

Software:

RPTrees is an implementation of the Random Projection Trees datastructure (here is the technical paper). It is a recursive space partitioning algorithm which has strong theoretical guarantee to adapt to the intrinsic structure of manifolds.

Useful Links:

Machine Learning blog
Sanjoy Dasgupta's webpage