Bharath Sriperumbudur

Bio

I am an associate professor in the Department of Statistics at the Pennsylvania State University.

My research interests include non-parametric statistics, machine learning, statistical learning theory, optimal transport and gradient flows, regularization and inverse problems, reproducing kernel spaces in probability and statistics, functional and topological data analysis.

My current research is supported by the award NSF-DMS-CAREER-1945396, "Statistical Learning, Inference and Approximation with Reproducing Kernels."

Publications

Preprints



Gromov-Wasserstein distances: Entropic regularization, duality and sample complexity

Z. Zhang, Z. Goldfeld, Y. Mroueh, and B. K. Sriperumbudur

[arxiv]



Spectral regularized kernel two-sample tests

O. Hagrass, B. K. Sriperumbudur, and B. Li

[arxiv]



Regularized Stein variational gradient flow

Y. He, K. Balasubramanian, B. K. Sriperumbudur, and J. Lu

[arxiv]



Shrinkage estimation of higher order Bochner integrals

S. Utpala, and B. K. Sriperumbudur

[arxiv]

 

Unified RKHS methodology and analysis for functional linear and single-index models

K. Balasubramanian, H-G. Muller, and B. K. Sriperumbudur

[arxiv]



Robust topological inference in the presence of outliers

S. Vishwanath, B. K. Sriperumbudur, K. Fukumizu and S. Kuruki

[arxiv]



Mean shrinkage estimation for high-dimensional diagonal natural exponential families

N. Siapoutis, D. Richards and B. K. Sriperumbudur

[arxiv]



On the limits of topological data analysis for statistical inference

S. Vishwanath, K. Fukumizu, S. Kuruki and B. K. Sriperumbudur

[arxiv]



Minimax estimation of quadratic Fourier functionals

S. Singh, B. K. Sriperumbudur and B. Poczos

[arxiv]



Gaussian processes and kernel methods: A review on connections and equivalences

M. Kanagawa, P. Hennig, D. Sejdinovic and B. K. Sriperumbudur

[arxiv]



Adaptive clustering using kernel density estimators

I. Steinwart, B. K. Sriperumbudur and P. Thomann

[arxiv]





2023



On distance and kernel measures of conditional dependence

T. Sheng and B. K. Sriperumbudur

Journal of Machine Learning Research, 24(7): 1-16, 2023. [pdf]



Optimal function-on-scalar regression over complex domains

M. Reimherr, B. K. Sriperumbudur and Hyun Bin Kang

Electronic Journal of Statistics, 17(1): 156-197, 2023. [pdf]





2022



Statistical optimality and computational efficiency of Nyström kernel PCA

N. Sterge and B. K. Sriperumbudur

Journal of Machine Learning Research, 23(337): 1-32, 2022. [pdf]



Approximate kernel PCA using random features: Computational vs. statistical trade-off

B. K. Sriperumbudur and N. Sterge

Annals of Statistics, 50(5): 2713-2736, 2022. [arxiv]



Cycle consistent probability divergences across different spaces

Z. Zhang, Y. Mroueh, Z. Goldfeld and B. K. Sriperumbudur

International Conference on Artificial Intelligence and Statistics, 2022. [arxiv]



Local minimax rates for closeness testing of discrete distributions

J. Lam-Weil, A. Carpentier and B. K. Sriperumbudur

Bernoulli, 28(2): 1179-1197, 2022. [arxiv]





2020



Robust persistence diagrams using reproducing kernels

S. Viswanath, K. Fukumzu, S. Kuruki and B. K. Sriperumbudur

Neural Information Processing Systems, 2020. [arxiv]



Gaussian sketching yields a J-L lemma in RKHS

S. Kpotufe and B. K. Sriperumbudur

International Conference on Artificial Intelligence and Statistics, 2020. [arxiv]



Gain with no pain: Efficient kernel-PCA by Nystrom sampling

N. Sterge, B. K. Sriperumbudur, L. Rosasco and A. Rudi

International Conference on Artificial Intelligence and Statistics, 2020. [arxiv]



Convergence analysis of deterministic kernel-based quadrature rules in misspecified settings

M. Kanagawa, B. K. Sriperumbudur and K. Fukumizu

Foundations of Computational Mathematics, 20, 155-194, 2020. [pdf]





2019



On kernel derivative approximation with random Fourier features

Z. Szabo and B. K. Sriperumbudur

International Conference on Artificial Intelligence and Statistics, 2019. [arxiv]





2018



Optimal prediction for additive function-on-function regression

M. Reimherr, B. K. Sriperumbudur and B. Taoufik

Electronic Journal of Statistics, 12(2), 4571-4601, 2018. [pdf]



Characteristic and universal tensor product kernels

Z. Szabo and B. K. Sriperumbudur

Journal of Machine Learning Research, 18(233): 1-29, 2018. [pdf]





2017



Minimax estimation of kernel mean embeddings

I. Tolstikhin, B. K. Sriperumbudur, and K. Muandet

Journal of Machine Learning Research, 18(86): 1-47, 2017. [pdf]



Density estimation in infinite dimensional exponential families

B. K. Sriperumbudur, K. Fukumizu, A. Gretton, A. Hyvarinen and R. Kumar

Journal of Machine Learning Research, 18(57):1-59, 2017. [pdf]



Kernel mean embedding of distributions: A review and beyond

K. Muandet, K. Fukumizu, B. K. Sriperumbudur and B. Scholkopf

Foundations and Trends in Machine Learning, 10(1-2):1-141, 2017. [arxiv]





2016



Convergence guarantees for kernel-based quadrature rules in misspecified settings

M. Kanagawa. B. K. Sriperumbudur and K. Fukumizu

Neural Information Processing Systems, 2016. [pdf]



Minimax estimation of maximal mean discrepancy with radial kernels

I. Tolstikhin, B. K. Sriperumbudur and B. Scholkopf

Neural Information Processing Systems, 2016. [pdf]



Learning theory for distribution regression

Z. Szabo, B. K. Sriperumbudur, B. Poczos and A. Gretton

Journal of Machine Learning Research, 17 (152):1-40, 2016. [pdf]



Kernel mean shrinkage estimators

K. Muandet, B. K. Sriperumbudur, K. Fukumizu, A. Gretton and B. Scholkopf

Journal of Machine Learning Research, 17 (48):1-41, 2016. [pdf]



On the optimal estimation of probability measures in weak and strong topologies

B. K. Sriperumbudur

Bernoulli, 22(3): 1839-1893, 2016. [arxiv]





2015



Optimal rates for random Fourier features

B. K. Sriperumbudur and Z. Szabo

Neural Information Processing Systems, 2015. [pdf]



Two-stage sampled learning theory on distributions

Z. Szabo, A. Gretton, B. Poczos and B. K. Sriperumbudur

International Conference on Artificial Intelligence and Statistics, 2015. [pdf]





2014



Kernel mean estimation via spectral filtering

K. Muandet, B. K. Sriperumbudur and B. Scholkopf

Neural Information Processing Systems, 2014. [pdf,supplement]



Kernel mean estimation and Stein's effect

K. Muandet, K. Fukumizu, B. K. Sriperumbudur, A. Gretton and B. Scholkopf

International Conference of Machine Learning, 2014. [pdf,supplement]





2013



Equivalence of distance-based and RKHS-based statistics in hypothesis testing

D. Sejdinovic, B. K. Sriperumbudur, A. Gretton and K. Fukumizu

Annals of Statistics, 41(5): 2263-2291, 2013. [pdf]

 

On the generalization ability of online learning algorithms for pairwise loss functions

P. Kar, B. K. Sriperumbudur, P. Jain and H. Karnick

International Conference on Machine Learning, 2013. [pdf]



Ultrahigh dimensional feature screening via RKHS embeddings

K. Balasubramanian, B. K. Sriperumbudur and G. Lebanon

International Conference on Artificial Intelligence and Statistics, 2013. [pdf,supplement]





2012



Optimal kernel choice for large-scale two-sample tests

A. Gretton, B. K. Sriperumbudur, D. Sejdinovic, H. Strathmann, S. Balakrishnan, M. Pontil and K. Fukumizu

Neural Information Processing Systems, 2012. [pdf]



On the empirical estimation of integral probability metrics

B. K. Sriperumbudur, K. Fukumizu, A. Gretton, B. Scholkopf and G. R. G. Lanckriet

Electronic Journal of Statistics, 6: 1550-1599, 2012. [pdf]



Hypothesis testing using pairwise distances and associated kernels

D. Sejdinovic, A. Gretton, B. K. Sriperumbudur and K. Fukumizu

International Conference on Machine Learning, 2012. [pdf]



A proof of convergence of the concave-convex procedure using Zangwill's theory

B. K. Sriperumbudur and G. R. G. Lanckriet

Neural Computation, 24(6): 1391–1407, 2012. [pdf]



Consistency and rates for clustering with DBSCAN

B. K. Sriperumbudur and I. Steinwart

International Conference on Artificial Intelligence and Statistics, 2012. [pdf,supplement]





2011

Learning in Hilbert vs. Banach spaces: A measure embedding viewpoint

B. K. Sriperumbudur, K. Fukumizu and G. R. G. Lanckriet

Neural Information Processing Systems, 2011. [pdf]

A majorization-minimization approach to the sparse generalized eigenvalue problem

B. K. Sriperumbudur, D. A. Torres and G. R. G. Lanckriet

Machine Learning, 85(1):3-39, 2011. [pdf]



Mixture density estimation via Hilbert space embedding of measures

B. K. Sriperumbudur

International Symposium on Information Theory, 2011. [pdf,slides]  

Universality, characteristic kernels and RKHS embedding of measures

B. K. Sriperumbudur, K. Fukumizu and G. R. G. Lanckriet

Journal of Machine Learning Research, 12(Jul): 2389-2410, 2011. [pdf]

2010

Reproducing kernel space embeddings and metrics on probability measures

B. K. Sriperumbudur

Ph. D. Dissertation, UC San Diego, 2010. [pdf]

Non-parametric estimation of integral probability metrics

B. K. Sriperumbudur, K. Fukumizu, A. Gretton, B. Scholkopf and G. R. G. Lanckriet

International Symposium on Information Theory, 2010. [pdf,slides]  

Hilbert space embeddings and metrics on probability measures

B. K. Sriperumbudur, A. Gretton, K. Fukumizu, B. Scholkopf and G. R. G. Lanckriet

Journal of Machine Learning Research, 11(Apr): 1297-1322, 2010. [pdf]

On the relation between universality, characteristic kernels and RKHS embedding of measures

B. K. Sriperumbudur, K. Fukumizu and G. R. G. Lanckriet

International Conference on Artificial Intelligence and Statistics, 2010. [pdf,slides]

2009

Kernel choice and classifiability for RKHS embeddings of probability distributions

B. K. Sriperumbudur, K. Fukumizu, A. Gretton, G. R. G. Lanckriet and B. Scholkopf

Neural Information Processing Systems, 2009. [pdf,slides]

Outstanding student paper award (Honorable mention)

On the convergence of the concave-convex procedure

B. K. Sriperumbudur and G. R. G. Lanckriet

Neural Information Processing Systems, 2009. [pdf]

2nd NIPS Workshop on Optimization for Machine Learning, 2009. [slides]

A fast, consistent kernel two-sample test

A. Gretton, K. Fukumizu, Z. Harchaoui and B. K. Sriperumbudur

Neural Information Processing Systems, 2009. [pdf,supplement]

Discussion of: Brownian distance covariance

A. Gretton, K. Fukumizu, and B. K. Sriperumbudur

Annals of Applied Statistics, 3(4): 1285-1294, 2009. [pdf]

A d.c. programming approach to the sparse generalized eigenvalue problem

B. K. Sriperumbudur, D. A. Torres and G. R. G. Lanckriet [arXiv]

2nd NIPS Workshop on Optimization for Machine Learning, 2009. [pdf,slides]

2008



Characteristic kernels on groups and semigroups

K. Fukumizu,  B. K. Sriperumbudur, A. Gretton and B. Scholkopf

Neural Information Processing Systems, 2008. [pdf]

RKHS representation of measures applied to homogeneity, independence and Fourier optics

B. Scholkopf, B. K. Sriperumbudur, A. Gretton and K. Fukumizu

Oberwolfach Report 30, Mathematisches Forschungsinstitut, Oberwolfach-Walke, Germany, pp. 42-44, 2008. [pdf]

Non-uniform speaker normalization using affine transformation

S. V. Bharath Kumar and S. Umesh

Journal of the Acoustical Society of America, 124(3), pp. 1727-1738, September 2008.

Injective Hilbert space embeddings of probability measures

B. K. Sriperumbudur, A. Gretton, K. Fukumizu, G. R. G. Lanckriet and B. Scholkopf

Conference on Learning Theory, 2008. [pdf,slides] 

Metric embedding for kernel classification rules

B. K. Sriperumbudur, O. Lang and G. R. G. Lanckriet

International Conference on Machine Learning, 2008. [pdf,slides]

2007

The effect of kernel choice on RKHS based statistical tests

B. K. Sriperumbudur, A. Gretton, K. Fukumizu and B. Scholkopf

Representations and Inference on Probability Distributions Workshop, NIPS 2007. [slides]

Finding musically meaningful words using sparse CCA

D. A. Torres, D. Turnbull, B. K. Sriperumbudur, L. Barrington and G. R. G. Lanckriet

Music, Brain and Cognition Workshop, NIPS 2007. [pdf,slides]

Sparse eigen methods by d.c. programming

B. K. Sriperumbudur, D. A. Torres and G. R. G. Lanckriet

International Conference on Machine Learning, 2007. [pdf,slides]



Nearest neighbor prototyping for sparse and scalable support vector       machines

B. K. Sriperumbudur and G. R. G. Lanckriet

Technical Report, Dept. of ECE, UCSD, February 2007. [pdf]

2006

Study of non-linear frequency warping functions for speaker normalization

S. V. Bharath Kumar, S. Umesh and Rohit Sinha

Proc. of IEEE International Conference on Acoustics, Speech and Signal Processing, 2006. [pdf]

A framework for parameter optimization in mutual information based registration algorithms

G. Gopalakrishnan, S. V. Bharath Kumar, A. Narayanan and R. Mullick

Proc. of SPIE Medical Imaging, 2006. [pdf]

2005

A fast piece-wise deformable method for multi-modality image registration

G. Gopalakrishnan, S. V. Bharath Kumar, A. Narayanan and R. Mullick

Proc. of Applied Imagery and Pattern Recognition, 2005. [pdf]

Lossless volumetric medical image compression with progressive multi-planar reformatting using 3-D DPCM

V. Nandedkar, S. V. Bharath Kumar and S. Mukhopadhyay

Proc. of National Conference on Image Processing, 2005. (Best Paper Award) [pdf]

Textural content in 3T MR: An image-based marker for Alzheimer's disease

S. V. Bharath Kumar, R. Mullick and U. Patil

Proc. of SPIE Medical Imaging, 2005. [pdf]

2004

Non-uniform speaker normalization using frequency-dependent scaling function

S. V. Bharath Kumar and S. Umesh

Proc. of International Conference on Signal Processing and Communications, 2004. [pdf]

A texture analysis approach for automatic flaw detection in pipelines

S. V. Bharath Kumar and S. Ramaswamy

Proc. of International Conference on Signal Processing and Communications, 2004. [pdf]

A novel progressive thick slab paradigm for volumetric medical image compression and navigation

S. V. Bharath Kumar, S. Mukhopadhyay and V. Nandedkar

Proc. of IEEE International Conference on Image Processing, 2004. [pdf]

Non-uniform speaker normalization using affine transformation

S. V. Bharath Kumar, S. Umesh and R. Sinha

Proc. of IEEE International Conference on Acoustics, Speech and Signal Processing, 2004. (rated amongst the top in its review category) [pdf]

An investigation into front-end signal processing for speaker normalization

S. Umesh, R. Sinha and S. V. Bharath Kumar

Proc. of IEEE International Conference on Acoustics, Speech and Signal Processing, Montreal, May 2004. [pdf]

Spatial distribution of T2 values in the hippocampus of Alzheimer's disease and control subjects

D. Blezek, S. V. Bharath Kumar, S. Adak, Z. Li, J. Schenck and E. Zimmerman

Twelfth ISMRM Scientific Meeting and Exhibition, 2004. (Poster)

2003

3-D loss-less multi-resolution image compression for medical images

S. Mukhopadhyay, S. V. Bharath Kumar, V. Nandedkar and A. Raparia

RSNA InfoRad presentation, 2003. (Poster)

Block-based conditional entropy coding for medical image compression

S. V. Bharath Kumar, N. Nagaraj, S. Mukhopadhyay and X. Xu

Proc. of SPIE Medical Imaging, 2003. [pdf]

2002

A simple approach to non-uniform vowel normalization

S. Umesh, S. V. Bharath Kumar, M. K. Vinay, R. Sharma and R. Sinha

Proc. of IEEE International Conference on Acoustics, Speech and Signal Processing, 2002. [pdf]

A model based approach to non-uniform vowel normalization

S. V. Bharath Kumar

M. Tech Thesis, Department of EE, IIT-K, 2002. [pdf]

1999

Realization of linear time-invariant system stability analyzers

S. V. Bharath Kumar

B. Tech Thesis, Department of ECE, SVU, Tirupati, 1999.