About me

I am a Research Scientist at Adobe Research since August 2017. I was a Postdoctoral Fellow in the School of Computer Science at Georgia Tech with Santosh Vempala during 2015-2017 where I also worked with Richard Peng. I obtained my PhD from the mathematics department at Yale University under the guidance of Daniel Spielman where I was also affiliated with Yale Institute For Network Science. Prior to that I received my bachelors degree in engineering physics from IIT Bombay.

Selected Publications (authors in alphabetical order unless marked *)

• David Durfee, Yu Gao, Anup B. Rao, Sebastin Wild. Efficient Second-Order Shape-Constrained Function Fitting. WADS'19
• *Thanh Tan Nguyen, Ali Shameli, Yasin Abbasi-Yadkori, Anup B. Rao, Branislav Kveton. Sample Efficient Graph-Based Optimization with Noisy Queries. AISTATS'19
• Michael B. Cohen, Jonathan Kelner,Rasmus Kyng, John Peebles, Richard Peng, Anup B Rao, Aaron Sidford, Solving Directed Laplacians in Nearly Linear Time through Sparse LU Factorizations. FOCS'18
• David Durfee, John Peebles, Richard Peng, Anup B. Rao, Determinant-Preserving Sparsification of SDDM Matrices with Applications to Counting and Sampling Spanning Trees. FOCS'17
• David Durfee, Rasmus Kyng, John Peebles, Anup B. Rao, Sushant Sachdeva. Sampling Random Spanning Trees Faster than Matrix Multiplication. (STOC 2017)
• Michael B. Cohen, Jonathan Kelner, John Peebles, Richard Peng, Anup B Rao, Aaron Sidford, Adrian Vladu. Almost-Linear-Time Algorithms for Markov Chains and New Spectral Primitives for Directed Graphs. (STOC 2017)
• Kevin Lai, Anup B. Rao, Santosh Vempala. Agnostic Estimation of Mean and Covariance. (FOCS 2016)
• Rasmus Kyng, Anup B. Rao, Sushant Sachdeva. Fast, Provable Algorithms for Isotonic Regression in all l_p-norms. (NIPS 2015)
• Rasmus Kyng, Anup B. Rao, Sushant Sachdeva, Daniel A. Spielman. Algorithms for Lipschitz Learning on Graphs. (COLT 2015)
• Peter Chin, Anup B. Rao and Van Vu. Stochastic Block Model and Community Detection in the Sparse Graphs: A spectral algorithm with optimal rate of recovery. (COLT 2015).
• Michael B. Cohen, Rasmus Kyng, Gary L. Miller, Jakub W. Pachocki, Richard Peng, Anup B. Rao and Shen Chen Xu. Solving SDD Linear Systems in Nearly mlog^1/2n Time. In STOC 2014. This paper is a merger of the following two results.
  • Michael B. Cohen, Rasmus Kyng, Jakub W. Pachocki, Richard Peng, and Anup B. Rao. Preconditioning in Expectation. (arXiv)
  • Michael B. Cohen, Gary L. Miller, Jakub W. Pachocki, Richard Peng, and Shen Chen Xu. Stretching Stretch. (arXiv)

Award

• Outstanding Post-Doctoral Research - From College of Computing in Georgia Tech.

Talks/Travels

• ICML, 2018
• Computational Challenges in Machine Learning, Simon's Institute, May 2017
• Simons Institute of ToC, Berkeley, March-May 2017
• UCSD, Feb 2017
• Simons Institute of ToC, Berkeley, Jan 2017
• Agnostic Estimation of Mean and Covariance, CMU, Jan 2017
• Graph Sparsification with Applications, Yale, Nov 2016
• Agnostic Estimation of Mean and Covariance,
  • FOCS, October 2016
  • University of Chicago, April 2016
  • Yale, April 2016
• Community Detection in Sparse Random Graphs
  • Allerton Conference, October 2015
  • SIAM, October 2015
  • ARC Colloquium
• Isotonic Regression
  • Machine Learning Seminar (Georgia Tech), January 2016
• Algorithms for Lipschitz Learning
  • ARC Colloquium, April 2015

Representative Teaching

• Spring 2016: Instructor (with Santosh Vempala), A Theoretician's Toolkit, Georgia Tech.
• Summer 2014: Mentor, Geometry of Polynomials, Summer Undergraduate Math Research at Yale, Yale University
• Fall 2013, Spring 2013, Fall 2011: Instructor, Calculus, Yale University

Code

• Robust Mean and Covariance Estimation: Our implementation of algorithms for estimating mean and covariance in the presence of outliers from Agnostic Estimation of Mean and Covariance.
• YINSlex Github Repository: Our implementations of the lex-minimization algorithms from the paper Algorithms for Lipschitz Learning on Graphs . The code was written by Rasmus Kyng, Sushant Sachdeva, Dan Spielman, and myself.
• Isotonic Github Repository: An implementation of the least-squares Isotonic regression algorithm from the paper Fast, Provable Algorithms for Isotonic Regression in all ℓpℓp-norms . The code was written by Rasmus Kyng, Sushant Sachdeva, and myself.