Publications

R-SPIDER: A Fast Riemannian Stochastic Optimization Algorithm with Curvature Independent Rate
Jingzhao Zhang, Hongyi Zhang, Suvrit Sra. (Nov 2018)
[.bib] [arXiv]

Finite sample expressive power of small-width ReLU networks
Chulhee Yun, Suvrit Sra, Ali Jadbabaie
[.bib] [arXiv]

Random Shuffling Beats SGD after Finite Epochs
Jefferey Z. HaoChen and Suvrit Sra.
[.bib] [arXiv]

On the computation of Wasserstein barycenters of multivariate Gaussians
Suvrit Sra.
[.bib] [arXiv]

Frank-Wolfe methods for geodesically convex optimization with application to the matrix geometric mean
Melanie Weber, Suvrit Sra.
[.bib] [arXiv]

Distributional Adversarial Networks
Chengtao Li, David Alvarez-Melis, Keyulu Xu, Stefanie Jegelka, Suvrit Sra. (May 2017)
[.bib] [arXiv] [Code]

@article{li2017distributional,
  title={Distributional Adversarial Networks},
  author={Chengtao Li, David Alvarez-Melis, Keyulu Xu, Stefanie Jegelka, Suvrit Sra},
  journal={arXiv preprint arXiv:1706.09549},
  year={2017}
}

Convex Optimization for Parallel Energy Minimization
K.S. Sesh Kumar, Álvaro J. Barbero, Suvrit Sra, Stefanie Jegelka, and Francis Bach.
[.bib] [arXiv]

@Article{sesh15,
 author = {K.S. Sesh Kumar and \'Alvaro Barbero and Stefanie Jegelka and  Suvrit Sra and Francis Bach},
 title = {{Convex Optimization for Parallel Energy Minimization}},
 journal = {arXiv:1503.01563},
 year = {2015},
}

An Alternative to EM for Gaussian Mixture Models: Batch and Stochastic Riemannian Optimization
Reshad Hosseini, Suvrit Sra. (Feb 2019)
Accepted Mathematical Programming, Series A
[.bib] [arXiv]

@Article{hoSr17,
 author = {Reshad Hosseini and Suvrit Sra},
 title = {{An Alternative to EM for Gaussian Mixture Models: Batch and Stochastic  Riemannian Optimization}},
 journal = {arXiv:1706.03267},
 year = {2017},
 note = {{\it Preprint}}
}

Small nonlinearities in activation functions create bad local minima in neural networks
Chulhee Yun, Suvrit Sra, and Ali Jadbabaie.
International Conference on Learning Representations (ICLR 2019)
[.bib] [arXiv]

Efficiently testing local optimality and escaping saddles for ReLU networks
Chulhee Yun, Suvrit Sra, and Ali Jadbabaie.
International Conference on Learning Representations (ICLR 2019)
[.bib] [arXiv]

Learning Determinantal Point Processes by Sampling Inferred Negatives
Zelda Mariet, Mike Gartrell, Suvrit Sra.
Artificial Intelligence and Statistics (AISTATS 2019)
[.bib] [arXiv]

Direct Runge-Kutta Discretization Achieves Acceleration
Jingzhao Zhang, Aryan Mokhtari, Ali Jadbabaie, and Suvrit Sra. (May 2018)
Advances in Neural Information Processing Systems (NIPS)
[.bib] [arXiv]

Exponentiated Strongly Rayleigh Measures
Zelda Mariet, Stefanie Jegelka, Suvrit Sra. (Dec 2018)
Advances in Neural Information Processing Systems (NIPS)
[.bib] [arXiv]

Modular proximal optimization with application to total variation regularization.
Álvaro J. Barbero, Suvrit Sra. submitted Nov 2013; (v2 Oct. 2014, v3 2016)
Journal of Machine Learning Research (JMLR)
[.bib] [arXiv] [.pdf]

@Article{barSra14,
 author = {\'Alvaro J. Barbero and Suvrit Sra},
 title = {{Modular proximal optimization for multidimensional total-variation regularization}},
 journal = {Journal of Machine Learning Research (JMLR)},
 year = {2018},
 note = {\it To appear (submitted: 2014; arXiv:1411.0589)}
}

New concavity and convexity results for symmetric polynomials and their ratios
Suvrit Sra. (Sep 2018)
Accepted: Linear and Multilinear Algebra
[.bib] [arXiv]

On Geodesically Convex Formulations for the Brascamp-Lieb Constant
Suvrit Sra, Nisheeth K. Vishnoi and Ozan Yıldız
21st International Conference on Approximation Algorithms for Combinatorial Optimization Problems (APPROX'2018)
[.bib] [arXiv]

An Estimate Sequence for Geodesically Convex Optimization
Hongyi Zhang, Suvrit Sra.
Conference on Learning Theory (COLT) 2018
[.bib] [arXiv]

Non-Linear Temporal Subspace Representations for Activity Recognition
Anoop Cherian, Suvrit Sra, Stephen Gould, Richard Hartley.
Computer Vision and Pattern Recognition (CVPR) 2018
[.bib] [arXiv]

@Article{chSrHa18,
 author = {Anoop Cherian and Suvrit Sra and Stephen Gould and Richard Hartley},
 title = {Non-Linear Temporal Subspace Representations for Activity Recognition},
 journal = {arXiv:1803.11064},
 year = {2018},
 note = {{\it CVPR 2018}}
}

Directional Statistics in Machine Learning: a Brief Review
Suvrit Sra.
In Applied directional statistics
[.bib] [.pdf]

Global optimality conditions for deep neural networks
Chulhee Yun, Suvrit Sra, Ali Jadbabaie.
International Conference on Learning Representations (ICLR 2018)
[.bib] [arXiv]

@Article{yuSrJa17,
author = {Chulhee Yun and Suvrit Sra and Ali Jadbabaie},
title = {Global optimality conditions for deep neural networks},
journal = {arXiv:1707:02444},
year = {2018},
note = {{\it ICLR 2018}}
}

A Generic Approach for Escaping Saddle points
S. Reddi, M. Zaheer, S. Sra, F. Bach, B. Poczos, R. Salakhutdinov, A. Smola
Artificial Intelligence and Statistics (AISTATS 2018)
[.bib] [arXiv]

Inequalities via symmetric polynomial majorization
Suvrit Sra
Accepted to Proceedings American Mathematical Society (PAMS) (Nov 2017)
[.bib] [arXiv] [.pdf]

@Article{sra15esym,
 author = {Suvrit Sra},
 title = {Inequalities via symmetric polynomial majorization},
 journal = {PAMS},
 year = {2015},
 note = {{\it arXiv:1509.05902}}
}

Elementary Symmetric Polynomials for Optimal Experimental Design
Zelda Mariet, Suvrit Sra
Advances in Neural Information Processing Systems (NIPS) 2017
[.bib] [arXiv]

Polynomial Time Dual Volume Sampling
Chengtao Li, Stefanie Jegelka, Suvrit Sra
Advances in Neural Information Processing Systems (NIPS) 2017
[.bib] [arXiv]

Combinatorial Topic Models using Small–Variance Asymptotics.
Ke Jiang, Suvrit Sra, Brian Kulis
Artificial Intelligence and Statistics (AISTATS) 2017
[.bib] [arXiv]

Kronecker Determinantal Point Processes
Zelda Mariet, Suvrit Sra
Advances in Neural Information Processing Systems (NIPS) 2016
[.bib] [arXiv]

Proximal Stochastic Methods for Nonsmooth Nonconvex Finite-Sum Optimization
Sashank Reddi, Suvrit Sra, Barnabas Poczos, Alex Smola
Advances in Neural Information Processing Systems (NIPS) 2016
[.bib] [.pdf] [arXiv]

Fast stochastic optimization on Riemannian manifolds
Hongyi Zhang, Sashank Reddi, Suvrit Sra
Advances in Neural Information Processing Systems (NIPS) 2016
[.bib] [arXiv]

Fast Mixing Markov Chains for Strongly Rayleigh Measures, DPPs, and Constrained Sampling
Chengtao Li, Stefanie Jegelka, Suvrit Sra
Advances in Neural Information Processing Systems (NIPS) 2016
[.bib] [arXiv]

Stochastic Frank-Wolfe Methods for Nonconvex Optimization
Sashank Reddi, Suvrit Sra, Barnabas Poczos, Alex Smola
54th Annual Allerton Conference on Communication, Control, and Computing. Dec 2016
[.bib] [.pdf] [arXiv]

Riemannian dictionary learning and sparse coding for positive definite matrices
Anoop Cherian, Suvrit Sra
IEEE Trans. Neural Networks and Learning Systems (TNNLS) 2016.
[.bib] [.pdf] [arXiv]

@Article{cheSra15,
 author = {Anoop Cherian and Suvrit Sra},
 title = {{Riemannian Dictionary Learning and Sparse Coding for Positive Definite Matrices}},
 journal = {arXiv:1507.02772},
 mon     = {Jul.},
 year = {2015},
}

Fast incremental method for smooth nonconvex optimization
Sashank J. Reddi, Suvrit Sra, Barnabas Poczos, Alexander J. Smola
IEEE Conference on Decision and Control (CDC), Dec 2016.
[.bib] [arXiv]

Positive Definite Matrices: Data Representation and Applications to Computer Vision
Anoop Cherian, Suvrit Sra.
Book chapter in Algorithmic Advances in Riemannian Geometry and Applications, Springer, 2016.
[.bib] [.pdf]

Geometric Optimization in Machine Learning
Suvrit Sra, Reshad Hosseini
Book chapter in Algorithmic Advances in Riemannian Geometry and Applications, Springer, 2016.
[.bib] [.pdf]

First-order methods for geodesically convex optimization
Hongyi Zhang, Suvrit Sra
Conference on Learning Theory (COLT 2016)
[.bib] [.pdf] [arXiv]

@Article{zhangSra16a,
 author = {Hongyi Zhang and Suvrit Sra},
 title = {First-order methods for geodesically convex optimization},
 journal = {arXiv:1602.06053},
 year = {2016},
 note = {{\it Preprint}}
}

Stochastic variance reduction for nonconvex optimization
Sashank Reddi, Ahmed Hefny, Suvrit Sra, Barnabas Poczos, Alexander J. Smola
International Conference on Machine Learning (ICML 2016)
[.bib] [.pdf] [arXiv]

@Article{reddi2016a,
 author = {Sashank Reddi and Ahmed Hefny and Suvrit Sra and Barnabas Poczos and Alexander J. Smola},
 title = {Stochastic variance reduction for nonconvex optimization},
 journal = {arXiv:1603.xxxx},
 year = {2016},
 note = {{\it Preprint}}
}

Geometric Mean Metric Learning.
Pourya H. Zadeh, Reshad Hosseini, Suvrit Sra
International Conference on Machine Learning (ICML 2016)
[.bib] [.pdf]

Fast DPP Sampling for Nyström with Application to Kernel Methods
Chengtao Li, Stefanie Jegelka, Suvrit Sra
International Conference on Machine Learning (ICML 2016)
[.bib] [.pdf] [arXiv]

Gaussian quadrature for matrix inverse forms with applications
Chengtao Li, Suvrit Sra, Stefanie Jegelka
International Conference on Machine Learning (ICML 2016)
[.bib] [.pdf] [arXiv]

Asynchronous Parallel Block-Coordinate Frank-Wolfe
Y.-X. Wang, V. Sadhanala, W. Dai, W. Neiswanger, Suvrit Sra, E. P. Xing
International Conference on Machine Learning (ICML 2016)
[.bib] [.pdf] [arXiv]

@Article{wangSaSra14,
 author = {Y.-X. Wang and V. Sadhanala and W. Dai and W. Neiswanger and Suvrit Sra and E. P. Xing},
 title = {{Asynchronous Parallel Block-Coordinate Frank-Wolfe}},
 journal = {arXiv:1409.6086},
 year = {2014},
}

On the Matrix Square Root via Geometric Optimization
Suvrit Sra
Electronic Journal of Linear Algebra (ELA) (May 2016).
[.bib] [arXiv] [.pdf]

@Article{sraRoot,
 author = {Suvrit Sra},
 title = {{On the matrix square root and geometric optimization}},
 journal = {arXiv:1507.08366},
 year = {2015},
 note = {\it Preprint}
}

Entropic Metric Alignment for Correspondence Problems
Justin Solomon, Gabriel Peyré, Vladimir Kim, Suvrit Sra
ACM SIGGRAPH 2016
[.bib] [.pdf] [supplement]

@Article{solomon16,
 author = {Justin Solomon and Gabriel Peyré and Vladimir Kim and Suvrit Sra},
 title = {{Entropic Metric Alignment for Correspondence Problems}},
 journal = {ACM SIGGRAPH},
 year = {2016},
}

Inference and mixture modelling with the Elliptical Gamma Distribution
Reshad Hosseini, Suvrit Sra, Lucas Theis, Matthias Bethge
Computational Statistics and Data Analysis (CSDA). Feb 2016
[.bib] [arXiv]

@Article{hoSra14,
 author = {Reshad Hosseini and Suvrit Sra and L. Theis and M. Bethge},
 title = {Statistical inference with the Elliptical Gamma Distribution},
 journal = {arXiv:1410.4812},
 year = {2014},
 note = {{\it Accepted to CSDA}},
}

Diversity Networks: Neural Network Compression Using Determinantal Point Processes
Zelda Mariet, Suvrit Sra
International Conference on Learning Representations (ICLR 2016)
[.bib] [arXiv]

@Article{marietSra15b,
 author = {Zelda Mariet and Suvrit Sra},
 title = {Diversity Networks},
 journal = {arXiv:1511.05077},
 year = {2015},
 note = {\it Preprint}
}

The sum of squared logarithms inequality in arbitrary dimensions
Lev Borisov, Patrizio Neff, Suvrit Sra, Christian Thiel
Linear Algebra and its Applications (LAA) Jan 2016.
[.bib] [arXiv]

@Article{boNes15,
 author = {Lev Borisov and Patrizio Neff and Suvrit Sra and Christian Thiel},
 title = {{The sum of squared logarithms inequality in arbitrary dimensions}},
 journal = {arXiv:1508.04039},
 year = {2015},
 note = {\it Preprint}
}

Efficient sampling for k-determinantal point processes
Chengtao Li, Stefanie Jegelka, Suvrit Sra
Artificial Intelligence and Statistics (AISTATS 2016)
[.bib] [arXiv]

@Article{ctli15,
 author = {Chengtao Li and Stefanie Jegelka and Suvrit Sra},
 title = {Efficient sampling for k-determinantal point processes},
 journal = {arXiv:1509.01618},
 year = {2015},
}

AdaDelay: Delay sensitive distributed stochastic convex optimization
Suvrit Sra, Adams Wei Yu, Mu Li, Alexander J. Smola
Artificial Intelligence and Statistics (AISTATS 2016)
[.bib] [arXiv]

@Article{sra15.distrib,
 author = {Suvrit Sra and Adams Wei Yu and Mu Li and Alexander J. Smola},
 title = {AdaDelay: Delay sensitive distributed stochastic convex optimization},
 journal = {arXiv:1508.05003},
 year = {2015},
 note = {\it Preprint}
}

Positive Definite Matrices and the S-Divergence
Suvrit Sra
Proceedings of the American Mathematical Society (PAMS). Oct 2015.
[.bib] [arXiv] [.pdf]

@Article{srasdiv,
 author = {Suvrit Sra},
 title = {{Positive Definite Matrices and the S-Divergence}},
 journal = {Proceedings of the American Mathematical Society},
 year = {2015},
 mon  = {Sep}
 note = {arXiv:1110.1773v4}
}

Matrix Manifold Optimization for Gaussian Mixtures
Reshad Hosseini, Suvrit Sra
Advances in Neural Information Processing Systems (NIPS 2015)
[.bib] [arXiv] .pdf]

@Article{hoSr15b,
 author = {Reshad Hosseini and Suvrit Sra},
 title = {{Manifold optimization for mixture modeling}},
 journal = {arXiv:1506.07677},
 year = {2015},
 note = {\it Submitted}
}

Asynchronous variance reduced stochastic gradient descent
Sashank Reddi, Ahmed Hefny, Suvrit Sra, Barnabas Poczos, Alexander J. Smola
Advances in Neural Information Processing Systems (NIPS 2015)
[.bib] [arXiv]

@Article{red15,
 author = {Sashank Reddy and Ahmed Hefny and Suvrit Sra and Barnabas Poczos and Alexander J. Smola},
 title = {Variance reduction in stochastic gradient},
 journal = {arXiv:1506.06840},
 year = {2015},
}

On inequalities for normalized Schur functions
Suvrit Sra
European Journal of Combinatorics. (submitted Feb 2015; accepted May 2015)
[.bib] [arXiv] [.pdf]

@Article{sra15a,
 author = {Suvrit Sra},
 title = {{On inequalities for normalized Schur functions}},
 journal = {European J. Combinatorics},
 volume = {Volume 51},
 year = {2016},
 pages = {492-–494},
 mon  = jan,
}

A proof of Thompson's determinantal inequality
Minghua Lin, Suvrit Sra
Mathematical Notes. (accepted Jun 2015).
[.bib] [arXiv]

@Article{linSra14,
 author = {Minghua Lin and Suvrit Sra},
 title = {{Complete strong superadditivity of generalized matrix functions}},
 journal = {Mathematical Notes},
 year = {2015},
}

Hlawka-Popoviciu inequalities on positive definite tensors
Wolfgang Berndt, Suvrit Sra
Linear Algebra and its Applications (accepted Feb 2015)
[.bib] [arXiv] [.pdf]

@Article{berSra15,
 author = {Wolfgang Berndt and Suvrit Sra},
 title = {{Hlawka-Popoviciu inequalities on positive definite tensors}},
 journal = {Linear Algebra and its Applications},
 volume = {486},
 number = {1},
 pages = {317--327},
 year = {2015},
 note = {arXiv:1411.0065}
}

Large-scale randomized-coordinate descent methods with non-separable linear constraints
Sashank J. Reddi, Ahmed Hefny, Carlton Downey, Abhinava Dubey, Suvrit Sra
Uncertainty in Artificial Intelligence (UAI 2015)
[.bib] [arXiv] [.pdf]

@Article{reHeSra14,
 author = {Sashank J. Reddi and Ahmed Hefny and Carlton Downey and Avinava Dubey and Suvrit Sra},
 title = {{Large-scale randomized-coordinate descent methods with non-separable linear constraints}},
 journal = {arXiv:1409.2617v3},
 mon     = {Oct.},
 year = {2014},
}

Fixed-point algorithms for learning determinantal point processes
Zelda Mariet, Suvrit Sra
International Conference on Machine Learning (ICML 2015)
[.bib] [arXiv] [.pdf]

@Inproceedings{marSra15,
 author = {Zelda Mariet and Suvrit Sra},
 title = {{Fixed-point algorithms for learning determinantal point processes}},
 booktitle = {International Conference on Machine Learning (ICML)},
 mon     = {Jun},
 year = {2015},
}

Conic geometric optimisation on the manifold of positive definite matrices
Suvrit Sra, Reshad Hosseini
SIAM Journal on Optimization (SIOPT) (accepted Jan 2015).
[.bib] [arXiv] [.pdf]

@Article{sraHo15,
 author = {Suvrit Sra and Reshad Hosseini},
 title = {{Conic Geometric Optimization on the Manifold of Positive Definite Matrices}},
 volume = {25},
 number = {1},
 pages  = {713--739},
 journal = {SIAM J. Optimization (SIOPT)},
 year = {2015},
}

Data Modeling with the Elliptical Gamma Distribution
Suvrit Sra, Reshad Hosseini, Lucas Theis, Matthias Bethge
Artificial Intelligence and Statistics (AISTATS 2015)
[.bib] [arXiv] [.pdf]

@InProceedings{hoSra15,
 author = {Reshad Hosseini and Suvrit Sra and L. Theis and M. Bethge},
 title = {Statistical inference with the Elliptical Gamma Distribution},
 booktitle = {Artificial Intelligence and Statistics (AISTATS)},
 year = {2015},
 volume = 18,
}

Efficient Structured Matrix Rank Minimization
Adams Wei Yu, Wanli Ma, Yaoliang Yu, Jaime G. Carbonell, Suvrit Sra
Advances in Neural Information Processing Systems (NIPS 2014)
[.bib] [arXiv] [.pdf]

@inproceedings{adams14,
 author    = {Adams Wei Yu and Wanli Ma and Yaoliang Yu and Jaime G. Carbonell and Suvrit Sra},
 title     = {Efficient Structured Matrix Rank Minimization},
 booktitle = {NIPS},
 year      = {2014},
 note      = {arXiv:1509.02447}
}
                

Riemannian sparse coding for positive definite matrices
Anoop Cherian, Suvrit Sra
European Conference on Computer Vision (ECCV 2014)
[.bib] [.pdf]

@inproceedings{chSra14,
  title={Riemannian sparse coding for positive definite matrices},
  author={Anoop Cherian and Suvrit Sra},
  booktitle={ECCV 2014},
  pages={299--314},
  year={2014},
  publisher={Springer}
}
                

Fast Newton methods for the group fused lasso
Matt Wytock, Suvrit Sra, Zico Kolter
Uncertainty in Artificial Intelligence (UAI 2014).
[.bib] [.pdf]

@inproceedings{wytock,
  title={Fast Newton methods for the group fused lasso},
  author={Matt Wytock and Suvrit Sra and Zico J. Kolter},
  booktitle={Proceedings of the 30th Conference on Uncertainty in Artificial Intelligence},
  year={2014}
}
                

Efficient nearest neighbors via robust sparse hashing
Anoop Cherian, Suvrit Sra, Vassilios Morellas, and Nikos Papanikolopoulos
IEEE Transactions on Image Processing, 23(8); 2014.
[.bib] [arXiv] [.pdf]

@article{chSra14b,
  title={Efficient nearest neighbors via robust sparse hashing},
  author={Anoop Cherian and Suvrit Sra and Vassilios Morellas and Nikolaos Papanikolopoulos},
  journal={IEEE Transactions on Image Processing},
  volume={23},
  number={8},
  pages={3646--3655},
  year={2014},
  publisher={IEEE}
}
                

Randomized Nonlinear Component Analysis
David Lopez-Paz, Suvrit Sra, Alexander J. Smola, Zoubin Ghahramani, and Bernhard Schölkopf
International Conference on Machine Learning (ICML 2014)
[.bib] [arXiv]

@Inproceedings{lopez,
  title={{Randomized Nonlinear Component Analysis}},
  author={David Lopez-Paz and Suvrit Sra and Alex Smola and Zoubin Ghahramani and Bernhard Schoelkopf},
  booktitle={Proceedings of the 31st International Conference on Machine Learning (ICML-14)},
  pages={1359--1367},
  year={2014}
}
                

Towards stochastic alternating direction method of multipliers
Samaneh Azadi and Suvrit Sra
International Conference on Machine Learning (ICML 2014)
[.bib] [.pdf]

@inproceedings{azadi,
  title={Towards an optimal stochastic alternating direction method of multipliers},
  author={Samaneh Azadi and Suvrit Sra},
  booktitle={Proceedings of the 31st International Conference on Machine Learning (ICML-14)},
  pages={620--628},
  year={2014}
}
                

Nonconvex proximal splitting: batch and incremental algorithms
Suvrit Sra
Invited book chapter: Regularization, Optimization, Kernels, and Support Vector Machines. (eds: J. A.K. Suykens, M. Signoretto, A. Argyriou. Mar 2014.
[.bib] [.pdf] [arXiv] [MPI-TR]

@InCollection{sra.ncprox,
  author =       {Suvrit Sra},
  title =        {Nonconvex proximal splitting: batch and incremental algorithms},
  booktitle =    {Regularization, Optimization, Kernels, and Support Vector Machines},
  publisher =    {Cambridge University Press},
  month     =    mar,
  year =         2014,
  editor =       {J. A. K. Suykens and M. Signoretto and A. Argyriou},
}
                

Tractable large-scale optimization in machine learning
Suvrit Sra
Invited book chapter: Tractability Practical Approaches to Hard Problems. (eds: L. Bordeaux, Y. Hamadi, P. Kohli); Aug 2013.
[.bib] [.pdf]

@InCollection{sra.optml,
  author =       {Suvrit Sra},
  title =        {Tractable Large-Scale Optimization in Machine Learning},
  booktitle =    {Advances in Tractability},
  publisher =    {Cambridge University Press},
  month     =    dec,
  year =         2013,
  editor =       {L. Bordeaux and Y. Hamadi and P. Kohli and R. Mateescu},
  note =         {29 pages},
}
                

Geometric optimisation on positive definite matrices with application to elliptically contoured distributions
Suvrit Sra and Reshad Hosseini
Advances in Neural Information Processing Systems (NIPS 2013)
[.bib] [arXiv] [.pdf]

@InProceedings{sraHo13,
  author =       {Suvrit Sra and Reshad Hosseini},
  title =        {{Geometric optimisation on positive definite matrices with application to elliptically contoured distributions}},
  booktitle =    {Advances in Neural Information Processing Systems (NIPS)},
  year =         2013,
  month =        dec,
}
                

Reflection methods for user-friendly submodular optimization
Stefanie Jegelka, Francis Bach, and Suvrit Sra
Advances in Neural Information Processing Systems (NIPS 2013)
[.bib] [arXiv] [.pdf]

@inproceedings{jeg13,
  title={Reflection methods for user-friendly submodular optimization},
  author={Stefanie Jegelka and Francis  Bach and Suvrit Sra},
  booktitle={Advances in Neural Information Processing Systems},
  pages={1313--1321},
  year={2013}
}
                

Correlation matrix nearness and completion under observation uncertainty
Carlos M. Alaiz, Francesco Dinuzzo, and Suvrit Sra
IMA Journal of Numerical Analysis Oct 2013.
[.bib] [.pdf] [preprint]

@Article{carlos,
  author =       {Carlos M. Ala\'iz and Francesco Dinuzzo and Suvrit Sra},
  title =        {Correlation matrix nearness and completion under observation uncertainty},
  journal =      {IMA Journal of Numerical Analysis},
  year =         {2013},
  month =        {Oct.},
  note =         {16 pages},
}
                

Jensen-Bregman LogDet Divergence for Efficient Similarity Computations on Positive Definite Tensors
Anoop Cherian, Suvrit Sra, Arindam Banerjee, and Nikos Papanikolopoulos
IEEE Transactions Pattern Analysis and Machine Intelligence (TPAMI) Dec. 2012.
[.bib] [.pdf]

@Article{chSra.pami,
  author =       {Anoop Cherian and Suvrit Sra and Arindam Banerjee and Nikolaos Papanikolopoulos},
  title =        {{Jensen-Bregman LogDet Divergence with Application to Efficient Similarity  Search for Covariance Matrices}},
  journal =      {IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)},
  year =         2012,
  month =        {Dec.},
  note =         {14 pages},
}
                

A new metric on the manifold of kernel matrices with application to matrix geometric means
Suvrit Sra
Advances of Neural Information Processing Systems (NIPS 2012)
[.bib] [.pdf]

@inproceedings{sra12.nips,
  title={A new metric on the manifold of kernel matrices with application to matrix geometric means},
  author={Suvrit Sra},
  booktitle={Advances in Neural Information Processing Systems},
  pages={144--152},
  year={2012}
}
                

Scalable nonconvex inexact proximal splitting
Suvrit Sra
Advances of Neural Information Processing Systems (NIPS 2012)
[.bib] [.pdf]

The multivariate Watson distribution: Maximum-likelihood estimation and other aspects
Suvrit Sra, Dmitrii B. Karp
Journal of Multivariate Analysis (accepted 2012)
[.bib] [arXiv] [.pdf] [.pdf]

Explicit eigenvalues of certain scaled trigonometric matrices
Suvrit Sra
Linear Algebra and its Applications (accepted Jul 2012)
[.bib] [arXiv] [.pdf]

Fast projection onto mixed-norm balls with applications
Suvrit Sra
Data Minining and Knowledge Discovery. 2012
[.bib] [arXiv] [.pdf]

A non-monotonic method for large-scale non-negative least squares
Dongmin Kim, Suvrit Sra, Inderjit S. Dhillon
Optimization Methods and Software (accepted: Dec. 2011)
[.bib] [.pdf]

Efficient Similarity Search for Covariance Matrices via the Jensen-Bregman LogDet Divergence
Anoop Cherian, Suvrit Sra, Arindam Banerjee, and Nikos Papanikolopoulos
International Conference on Computer Vision (ICCV) (2011)
[.bib] [.pdf]; [Bugfix .pdf]

Generalized Dictionary Learning for Symmetric Positive Definite Matrices with Application to Nearest Neighbor Retrieval
Suvrit Sra, Anoop Cherian
European Conf. on Machine Learning (ECML) (2011)
[.bib] [.pdf]

Fast projections onto L1,q-norm balls for grouped feature selection
Suvrit Sra
European Conference on Machine Learning (ECML 2011). Best Paper Runner Up
[.bib] [.pdf]

Fast Newton-type Methods for Total-Variation with Applications
Álvaro J. Barbero, Suvrit Sra
International Conference on Machine Learning (ICML 2011)
[.bib] [.pdf]

A short note on parameter approximation for von Mises-Fisher distributions: and a fast implementation of I_s(x)
Suvrit Sra
Computational Statistics 2011.
[.bib] [.pdf]

Optimization for Machine Learning
Suvrit Sra, Sebastian Nowozin, Stephen J. Wright
MIT Press, 2011.
[MIT Press] [Amazon] [Barnes and Noble]

Projected Newton-type methods in machine learning
Mark Schmidt, Dongmin Kim, Suvrit Sra
In: "Optimization for Machine Learning": MIT Press, 2011.
[.bib] [.pdf]

Online Multi-frame Blind Deconvolution with Super-resolution and Saturation Correction
Michael Hirsch, Stefan Harmeling, Suvrit Sra, Bernhard Schölkopf
Astronomy & Astrophysics Feb (2011)
[.bib] [.pdf]

Denoising sparse noise via online dictionary learning
Anoop Cherian, Suvrit Sra, Nikos Papanikolopoulos
IEEE Conference on Speech Acoustics and Signal Processing (ICASSP 2011)
[.bib] [.pdf]

Sparse inverse covariance estimation using an adaptive gradient method
Suvrit Sra and Dongmin Kim
[.bib] [arXiv]

Tackling box-constrained convex optimization via a new projected quasi-Newton approach
Dongmin Kim, Suvrit Sra, Inderjit S. Dhillon
SIAM Journal on Scientific Computing (SISC). Oct 2010
[.bib] [.pdf

A scalable trust-region algorithm with application to mixed-norm regression
Dongmin Kim, Suvrit Sra, Inderjit S. Dhillon
Interational Conference on Machine Learning (ICML 2010)
[.bib] [.pdf]

Multiframe Blind Deconvolution, Super-Resolution, and Saturation Correction via Incremental EM
Stefan Harmeling, Suvrit Sra, Michael Hirsch, Bernhard Schölkopf
IEEE International Conference on Image Processing (ICIP 2010)
[.bib] [.pdf]

Efficient Filter Flow for Space-Variant Multiframe Blind Deconvolution
Michael Hirsch, Suvrit Sra, Bernhard Schölkopf, Stefan Harmeling
IEEE Conference Computer Vision & Pattern Recognition (CVPR 2010)
[.bib] [.pdf]

Sparse nonnegative matrix approximation: new formulations and algorithms
Rashish Tandon and Suvrit Sra
MPI Technical Report #193. Sep 2010
[.bib] [.pdf]

Fast algorithms for total-variation based optimization
Alvaro J. Barbero and Suvrit Sra
MPI Technical Report #194. Aug 2010
[.bib] [.pdf]

Generalized proximity and projection with norms and mixed-norms
Suvrit Sra
MPI Technical Report #192. May 2010
[.bib] [.pdf]

Efficient Filter Flow for Space-Variant Multiframe Blind Deconvolution
Michael Hirsch, Suvrit Sra, Bernhard Schölkopf and Stefan Harmeling
MPI Technical Report #188 Nov 2009
[.bib] [.pdf]

Text Clustering with Mixture of von Mises-Fisher Distributions
Arindam Banerjee, Inderjit S. Dhillon, Joydeep Ghosh, Suvrit Sra
In: "Text Mining: Theory, Applications, and Visualization" eds. A. N. Srivastava and M. Sahami, CRC Press. 2009.
[.bib] [.pdf]

Approximation Algorithms for Tensor clustering
Stefanie Jegelka, Suvrit Sra, Arindam Banerjee
Algorithmic Learning Theory (ALT 2009).
[.bib] [arXiv]

Online Blind Deconvolution for Astronomy
Stefan Harmeling, Michael Hirsch, Suvrit Sra, Bernhard Schölkopf
IEEE Interational Conferemce on Computational Photography (ICCP 2009)
[.bib] [.pdf]

A new non-monotonic algorithm for PET image reconstruction
Suvrit Sra, Dongmin Kim, Inderjit S. Dhillon, Bernhard Schölkopf
IEEE Nuclear Science Symposium / Medical Imaging Conf. (NSS/MIC 2009)
[.bib] [Conference Record M03-2: ]

Scalable Semidefinite Programming using Convex Perturbations
Brian Kulis, Suvrit Sra, Inderjit S. Dhillon
Artificial Intelligence and Statistics (AISTATS 2009)
[.bib] [.pdf]

Block-Iterative Algorithms for Non-negative Matrix Approximation
Suvrit Sra
IEEE International Conference on Data Mining (ICDM 2008)
[.bib] [.pdf]

The Metric Nearness Problem
Justin Brickell, Inderjit S. Dhillon, Suvrit Sra, Joel A. Tropp
>SIAM Journal on Matrix Analysis and Applications (SIMAX). 30(1). pp. 375--396 (2008).
SIAM Outstanding Paper Prize 2011 -- across SIAM Journals in the three years 2008--2010
[.bib] [.pdf]

Approximation Algorithms for Bregman Clustering Co-clustering and Tensor Clustering
Suvrit Sra, Stefanie Jegelka, and Arindam Banerjee
MPI Technical Report #177 2008.
[.bib] [.pdf]

Block iterative algorithms for non-negative matrix approximation
Suvrit Sra
MPI Technical Report #176 2008
[.bib] [.pdf]

Non-monotonic Poisson Likelihood Maximization
Suvrit Sra, Dongmin Kim, and Bernhard Schölkopf
MPI Technical Report #170. Jun 2008
[.bib] [.pdf]

A New Non-monotonic Gradient Projection Method for the Non-negative Least Squares Problem
Dongmin Kim, Suvrit Sra, and Inderjit S. Dhillon
Computer Sciences, University of Texas at Austin, TR-08-28.
[.bib] [.pdf]

Matrix Nearness Problems in Data Mining
Suvrit Sra
Ph.D. Thesis. University of Texas at Austin. Aug. 2007
Thesis: .pdf [.bib] [arXiv] [.pdf]

Information-theoretic Metric Learning
Jason V. Davis, Brian Kulis, Prateek Jain, Suvrit Sra, Inderjit S. Dhillon
International Conference on Machine Learning (ICML) 2007.; (Best Student Paper)
Paper: .pdf [.bib] [arXiv] [.pdf]

Fast Newton-type Methods for the Least Squares Nonnegative Matrix Approximation Problem
Dongmin Kim, Suvrit Sra, Inderjit S. Dhillon
SIAM International Conference on Data Mining (SDM) 2007; Best of SDM 2007 papers
Paper: .pdf [.bib] [arXiv] [.pdf]

Fast Projection-Based Methods for the Least Squares Nonnegative Matrix Approximation Problem
Dongmin Kim, Suvrit Sra, Inderjit S. Dhillon
in Statistical Analysis and Data Mining vol. 1 pp. 38-51 (2007);
[.pdf]; [author PDF] [.bib]

Scalable Semidefinite Programming using Convex Perturbations
Brian Kulis, Suvrit Sra, Stefanie Jegelka, and Inderjit S. Dhillon
Comp. Sci., Univ. of Texas at Austin, TR-07-47, Sep. 2007;
Paper: .pdf [.bib] [arXiv] [.pdf]

A New Projected Quasi-Newton Approach for solving the Nonnegative Least-Squares Problem
Dongmin Kim, Suvrit Sra, and Inderjit S. Dhillon
Comp. Sci., Univ. of Texas at Austin, TR-06-54, May 2007;
Paper: .pdf [.bib] [arXiv] [.pdf]

Modeling data using directional distributions: Part II
Suvrit Sra, Prateek Jain, and Inderjit S. Dhillon
Comp. Sci., Univ. of Texas at Austin, TR-07-05, Feb. 2007;
Paper: .pdf [.bib] [arXiv] [.pdf]

Information-theoretic Metric Learning
Jason V. Davis, Brian Kulis, Suvrit Sra, and Inderjit S. Dhillon
NIPS 2006 Workshop on learning to compare examples, Dec. 2006;
Paper: .pdf [.bib] [arXiv] [.pdf]

Incremental Aspect Models for Mining Document Streams
Arun Surendran, Suvrit Sra
in Principles and Practice of Knowledge Discovery in Databases (PKDD) 2006;
Paper: [.pdf [.bib] [arXiv] [.pdf]

Efficient Large Scale Linear Programming Support Vector Machines
Suvrit Sra
in European Conference on Machine Learning (ECML) 2006.;
Paper: [.pdf [.bib] [arXiv] [.pdf]

Row-action Methods for Compressed Sensing
Suvrit Sra, Joel A. Tropp
in International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 2006.;
Paper: [.pdf [.bib] [arXiv] [.pdf]

Clustering on the Unit Hypersphere using von Mises-Fisher Distributions
Arindam Banerjee, Inderjit S. Dhillon, Joydeep Ghosh, Suvrit Sra
in Journal of Machine Learning Research (JMLR) vol. 6 pp. 1345-1382 (2005);
Paper: [.pdf [.bib] [arXiv] [.pdf]

Generalized Nonnegative Matrix Approximations with Bregman Divergences
Inderjit S. Dhillon, Suvrit Sra
in Advances Neural Information Processing Systems (NIPS) 2005.;
Paper: [.pdf [.bib] [arXiv] [.pdf]

Minimum Sum Squared Residue based Co-clustering of Gene Expression data
Hyuk Cho, Inderjit S. Dhillon, Yuqiang Guan, Suvrit Sra
in SIAM International Conference on Data Mining (SDM) 2004;
Paper: [.pdf [.bib] [arXiv] [.pdf]

Triangle Fixing Algorithms for the Metric Nearness Problem
Inderjit S. Dhillon, Suvrit Sra, J. A. Tropp
Advances in Neural Information Processing Systems (NIPS) 2004;
Paper: [.pdf [.bib] [arXiv] [.pdf]

Generative Model-Based Clustering of Directional Data
Arindam Banerjee, Inderjit S. Dhillon, Joydeep Ghosh, Suvrit Sra
International Conference on Knowledge Discovery and Data Mining (KDD) 2003.;
Paper: [.pdf [.bib] [arXiv] [.pdf]

Nonnegative Matrix Approximation: Algorithms and Applications
Suvrit Sra and Inderjit S. Dhillon
Comp. Sci., Univ. of Texas at Austin TR-06-27, Jun 2006;
Paper: .pdf [.bib] [arXiv] [.pdf]

Generalized Nonnegative Matrix Approximations using Bregman Divergences
Inderjit S. Dhillon and Suvrit Sra
Comp. Sci., Univ. of Texas at Austin TR-05-31, Jun 2005;
Paper: .pdf [.bib] [arXiv] [.pdf]

Triangle Fixing Algorithms for the Metric Nearness Problem
Inderjit S. Dhillon, Suvrit Sra, and Joel A. Tropp
Comp. Sci., Univ. of Texas at Austin TR-04-22, Jun 2004;
Paper: .pdf [.bib] [arXiv] [.pdf]

The Metric Nearness Problem with Applications
Inderjit S. Dhillon, Suvrit Sra, and Joel A. Tropp
Comp. Sci., Univ. of Texas at Austin TR-03-23, July 2003;
Paper: .ps.gz [.bib] [arXiv] [.pdf]

Expectation Maximization for Clustering on Hyperspheres
Arindam Banerjee, Inderjit S. Dhillon, Joydeep Ghosh, and Suvrit Sra
Comp. Sci., Univ. of Texas at Austin TR-03-07, Feb. 2003;
Paper: .ps.gz [.bib] [arXiv] [.pdf]

Modeling Data using Directional Distributions
Inderjit S. Dhillon and Suvrit Sra
Comp. Sci., Univ. of Texas at Austin TR-03-06, Jan. 2003;
Paper: .ps.gz [.bib] [arXiv] [.pdf]