(for the very latest, see my arXiv papers, google scholar papers, DBLP computer science papers, or ResearchGate and academia.edu profiles; my ORCID)
Global solutions to deep learning in sub-linear time, Stephen Becker
Locality-sensitive hashing in function spaces, Will Shand and Stephen Becker
Guarantees for the Kronecker Fast Johnson-Lindenstrauss Transform Using a Coherence and Sampling Argument, Osman Malik and Stephen Becker
Resolvability of Hamming Graphs, Lucas Laird, Richard C. Tillquist, Stephen Becker, Manuel E. Lladser
Randomization of Approximate Bilinear Computation for Matrix Multiplication, Osman Malik and Stephen Becker
Stochastic Subspace Descent, David Kozak, Stephen Becker, Alireza Doostan, Luis Tenorio
Penalized basis models for very large spatial datasets, Mitchell Krock, William Kleiber, Stephen Becker
Tensor Robust Principal Component Analysis: Better recovery with atomic norm regularization, Derek Driggs, Stephen Becker and Jordan Boyd-Graber
Fast Randomized Matrix and Tensor Interpolative Decomposition Using CountSketch, Osman Malik and Stephen Becker. Code on github
Efficient Solvers for Sparse Subspace Clustering, Farhad Pourkamali-Anaraki, Stephen Becker
URV Factorization with Random Orthogonal System Mixing, Stephen Becker, James Folberth, Laura Grigori
Optimization and Learning with Information Streams: Time-varying Algorithms and Applications, Emiliano Dall'Anese, Andrea Simonetto, Stephen Becker, Liam Madden, to appear May 2020 in IEEE Signal Processing Magazine
Safe Feature Elimination for Non-Negativity Constrained Convex Optimization, James Folberth, Stephen Becker. To appear in J. Optimization Theory and Applications (JOTA), 2019. Here is the post-peer-review, pre-copyedit version of the article that is published in JOTA. The final authenticated version is available online at dx.doi.org/10.1007/s10957-019-01612-w or the free view-only version.
Improved Fixed-Rank Nystrom Approximation via QR Decomposition: Practical and Theoretical Aspects, Farhad Pourkamali-Anaraki, Stephen Becker. Neurocomputing 363(21) pp 261–272, 2019 DOI
Stochastic Lanczos estimation of genomic variance components for linear mixed-effects models, Richard Border and Stephen Becker. BMC Bioinformatics, final version (open access), 2019
On Quasi-Newton Forward–Backward Splitting: Proximal Calculus and Convergence, Stephen Becker, Jalal Fadili, Peter Ochs. arXiv link. SIAM J. Optimization, 29(4), 2445-2482, 2019. DOI link
Adapting Regularized Low Rank Models for Parallel Architectures, Derek Driggs, Stephen Becker, Aleksandr Aravkin. SIAM J. Sci. Computing, 2019 (DOI link)
Template Polyhedra and Bilinear Optimization, Jessica Gronski, Mohamed-Amin Ben Sassi, Stephen Becker, Sriram Sankaranarayanan. Formal Methods and Design, 2018.
Achieving superresolution with illumination-enhanced sparsity, Jiun-Yann Yu, Stephen R. Becker, James Folberth, Bruce F. Wallin, Simeng Chen, and Carol J. Cogswell. Optics Express, vol 26(8) pp. 9850-9865 (2018) . Researchgate full text
Preconditioned Data Sparsification for Big Data with Applications to PCA and K-means, F. Pourkamali-Anaraki, S. Becker. IEEE Trans. Info Theory, 2017, DOI link. Also available via CU Scholar website.
Efficient Adjoint Computation for Wavelet and Convolution Operators (Lecture Notes), J. Folberth and S. Becker, IEEE Signal Processing Magazine, vol 33(6) 135–147, Nov 2016 Official link or arXiv
Designing Statistical Estimators That Balance Sample Size, Risk, and Computational Cost, J. Bruer, J. Tropp, V. Cevher and S. Becker, IEEE Journal of Selected Topics in Signal Processing link, Vol 9, No 4, June 2015
Convex Optimization for Big Data: Scalable, randomized, and parallel algorithms for big data analytics, Volkan Cevher, Stephen Becker, Mark Schmidt, IEEE Signal Processing Magazine, vol. 31, no. 5, 2014. arXiv version also available
An Algorithm for Splitting Parallel Sums of Linearly Composed Monotone Operators, with Applications to Signal Recovery, Stephen Becker and Patrick L. Combettes, Journal of Nonlinear and Convex Analysis, vol. 15, no. 1, pp. 137–159, Jan. 2014
Improving IMRT delivery efficiency with iteratively reweighted L1-minimization for inverse planning, with Hojin Kim, Lei Xing, Ruijiang Li et al., Medical Physics, vol. 40, 071719, 2013. DOI
A Non-Uniform Sampler for Wideband Spectrally-Sparse Environments, Michael Wakin, Stephen Becker, Eric Nakamura, Michael Grant, Emilio Sovero, Daniel Ching, Juhwan Yoo, Justin Romberg, Azita Emami-Neyestanak, Emmanuel Candès, IEEE Journal on Emerging and Selected Topics in Circuits and Systems (JETCAS), vol. 2, no. 3, pp. 516–529, 2012. DOI
A Compressed Sensing Parameter Extraction Platform for Radar Pulse Signal Acquisition, J. Yoo, C. Turnes, E. Nakamura, C. Le, S. Becker, E. Sovero, M. Wakin, M. Grant, J. Romberg, A. Emami-Neyestanak, and E. Candès, IEEE Journal on Emerging and Selected Topics in Circuits and Systems (JETCAS), vol. 2, no. 3, pp. 626–638, 2012. DOI
Dynamical Behavior Near a Liquid-Liquid Phase Transition in Simulations of Supercooled Water by P. H. Poole, S. R. Becker, F. Sciortino, and F. W. Starr; Journal of Physical Chemistry B, vol. 114, no. 48, pp. 14176–14183, 2011.
Templates for Convex Cone Problems with Applications to Sparse Signal Recovery by S. Becker, E. Candès, M. Grant. Mathematical Programming Computation, vol 3 no 3 (2011) 165–218. Software. An extended version appears as chapter 4 in my thesis. Typos: eq. 4.6a should have 1, not sigma_k^{-1}t_k^{(1)}, in the second parameter of Trunc, and the two equations above should have z^{(1)} not z^{(2)} in the l-infinity norm constraint.
NESTA: a fast and accurate first-order method for sparse recovery by S. Becker, J. Bobin and E. J. Candès. SIAM J. on Imaging Sciences (SIIMS), vol 4 no. 1 (2011). Software. An extended version appears as chapter 2 in my thesis.
Quantum state tomography via compressed sensing by Gross, Liu, Flammia, Becker and Eisert, in Physical Review Letters, 2010, vol 105, number 15 (4 pages).
Relation between the Widom line and the breakdown of the Stokes-Einstein relation in supercooled water, P. Kumar, S.V. Buldyrev, S.R. Becker, P.H. Poole, F.W. Starr, and H.E. Stanley, Proceedings of the National Academy of Science, Vol. 104, 9575-9579 (2007).
Fractional Stokes-Einstein and Debye-Stokes-Einstein Relations in a Network-Forming Liquid, by Stephen R. Becker, Peter H. Poole, and Francis W. Starr, in Physical Review Letters, vol. 97 no. 5. Copyright 2006 by the American Physical Society.
“The Dynamics of Falling Dominoes”, Stan Wagon, Adrianne Pontarelli, Stephen Becker and William Briggs, UMAP Journal, (26)1, 2005, pp. 37-48.
Dual Smoothing Techniques for Variational Matrix Decomposition, S. Becker and A. Aravkin, in “Robust Low-Rank and Sparse Matrix Decomposition: Applications in Image and Video Processing”, T. Bouwmans, N. Aybat, E. Zahzah, eds. CRC Press, 2016. arXiv link
One-Pass Sparsified Gaussian Mixtures, Eric Kightley and Stephen Becker, to appear in the Machine Learning workshop, at the 2019 IEEE International Conference on Big Data (BigData 2019), Dec 9-12, 2019, Los Angeles, CA, USA (“special session” on Machine Learning in Big Data). Paper available on arXiv.
Online Sparse Subspace Clustering, Liam Madden, Stephen Becker, Emiliano Dall'Anese (in the 2nd IEEE Data Science Workshop, June 2019 Minneapolis). DOI link
Perturbed Proximal Descent to Escape Saddle Points for Non-convex and Non-smooth Objective Functions, Zhishen (Leo) Huang and Stephen Becker, in INNS Big Data and Deep Learning 2019 (Sestri Levante, Genova, Italy 16-18 April 2019). Published version appears in Recent Advances in Big Data and Deep Learning. DOI link
Low-rank Tucker decomposition of large tensors using TensorSketch, S. Becker and O. Malik, in NIPS 2018 (aka NeurIPS 2018). Code on github
Randomized Clustered Nystrom for Large-Scale Kernel Machines, F. Pourkamali-Anaraki, S. Becker, M. Wakin. Conference on Artificial Intelligence (AAAI 2018), San Francisco, Feb 2018. (A preliminary, shorter version appears as Randomized Clustered Nystrom for Large-Scale Kernel Machines, Farhad Pourkamali-Anaraki, Stephen Becker )
Estimating Active Subspaces with Randomized Gradient Sampling, S Becker and F. Pourkamali-Anaraki, SIAM Workshop on Dimension Reduction 2017, July 9-10, Pittsburg. Also available via CU Scholar website.
Robust Partially-Compressed Least-Squares, S. Becker, B. Kawas, M. Petrik, K. N. Ramamurthy. In AAAI 2017
A Randomized Approach to Efficient Kernel Clustering, F. Pourkamali-Anaraki, S. Becker, in 2016 IEEE GlobalSIP (Global Conference on Signal and Information Processing). Also available via CU Scholar website.
Efficient Dictionary Learning via Very Sparse Random Projections, F. Pourkamali-Anaraki, S. Becker, S. Hughes, in SampTA (May 2015). DOI link
General Optimization Framework for Robust and Regularized 3D FWI, S. Becker, L. Horesh, A. Aravkin, E. van den Berg, S. Zhuk, as extended abstract at EAGE (Madrid, Spain, June 2015)
Time–Data Tradeoffs by Aggressive Smoothing, J. Bruer, J. A. Tropp, V. Cevher, S. Becker, NIPS 2014, Montreal.
QUIC & DIRTY: A Quadratic Approximation Approach for Dirty Statistical Models, C-J. Hsieh, I. Dhillon, P. Ravikumar, S. Becker, P. Olsen, NIPS 2014, Montreal.
A variational approach to stable principal component pursuit, A. Aravkin, S. Becker, V. Cevher, P. Olsen, UAI 2014. PDF
Joint low-rank representation and matrix completion under a singular value thresholding framework, C. Tzagkarakis, S. Becker and A. Mouchtaris, EUSIPCO 2014.
Metric learning with rank and sparsity constraints, B. Bah, S. Becker, V. Cevher and B. Gozcu, ICASSP 2014.
A proximal splitting method for inf-convolutive variational models in image recovery, S. Becker and P. L. Combettes, accepted in ICIP 2013, Melbourne
Scalable and accurate quantum tomography from fewer measurements, Stephen Becker and Volkan Cevher, SPARS 2013, Lausanne Switzerland.
Randomized Singular Value Projection, Stephen Becker, Volkan Cevher, Anastasios Kyrillidis, SampTA 2013, Bremen Germany. This arXiv version is an extended version (i.e., with proofs) of the 4 page SampTA proceedings version.
Sparse projections onto the simplex, Stephen Becker, Volkan Cevher, Christopher Koch, Anastasios Kyrillidis (the v3 on arXiv is our expanded version of our NIPS conference version). ICML, Atlanta. For fun, see the ICML 2013 review boards. There is an earlier version on the JMLR website.
Sparse projections onto the simplex, Anastasios Kyrillidis, Stephen Becker, Volkan Cevher, in NIPS Workshop on Discrete Optimization in Machine Learning (DISCML), Dec. 2012
A quasi-Newton proximal splitting method, Stephen Becker, M. Jalal Fadili, in NIPS 2012 (Lake Tahoe) (recieved a “spotlight presentation” as one of the top 5% of submissions)
A 100MHz-2GHz 12.5x sub-Nyquist Rate Receiver in 90nm CMOS, J. Yoo, S. Becker, M. Loh, M. Monge, E. Candès, A. Emami-Neyestanak. In RFIC 2012 (Montreal, Canada, June 2012). DOI
Design and implementation of a fully integrated compressed-sensing signal acquisition system, J. Yoo, S. Becker, M. Monge, M. Loh, E. Candès, A. Emami-Neyestanak. In ICASSP 2012 (Kyoto, Japan, March 2012) DOI
(full list of invited talks is on my CV; here I list a few that have video or slides or websites)
“Matrix Completion and Robust PCA”, University of Colorado Boulder, Computer Science department colloquium, Boulder, CO. Nov 20 2014. Video available here
“Advances in first-order methods: constraints, non-smoothness and faster convergence”, Minisymposium, SIAM Imaging Science, Philadelphia. Slides, May 22 2012
“TFOCS: Flexible First-order Methods for Rank Minimization”, at the Low-rank Matrix Optimization minisymposium at the 2011 SIAM Conference on Optimization in Darmstadt. Here's the RPCA video (Matlab code to generate this can be found at the Demo page on the TFOCS website or at the locally hosted version of the RPCA demo), May 19 2011
“Quick intro to convex optimization” talk for ACM^tea, Oct 23 2009
The Chen-Teboulle algorithm is the proximal point algorithm, from 2011 but posted 2019, shows the Chen-Teboulle algorithm admits a more aggressive stepsize than via the original analysis.
Exact linesearch for LASSO discusses exact step-size selection for piecewise quadratic objective functions, with code at github
Practical Compressed Sensing: modern data acquisition and signal processing, California Institute of Technology (PhD thesis). Co-winner, Carey prize.
Translational and Rotational Dynamics of Supercooled Water, Wesleyan University (undergraduate thesis). Co-winner, Bertman prize.
Theses of my students:
A comparison of spectral estimation techniques, Marc Thomson, 2019, Masters; code at github.
Stokes, Gauss, and Bayes walk into a bar…, Eric Kightley, 2019, PhD
Non-Convex Optimization and Applications to Bilinear Programming and Super-Resolution Imaging, Jessica Gronski, 2019, PhD
Stochastic Lanczos Likelihood Estimation of Genomic Variance Components, Richard Border, 2018, Masters
Fast and Reliable Methods in Numerical Linear Algebra, Signal Processing, and Image Processing, James Folberth, 2018, PhD
Randomized Algorithms for Large-Scale Data Analysis, Farhad Pourkamali Anaraki, 2017, PhD
Optimization for High-Dimensional Data Analysis, Derek Driggs, 2017, Masters
Open Problem: Property Elicitation and Elicitation Complexity, Rafael Frongillo, Ian Kash, Stephen Becker. 29th Annual Conference on Learning Theory (COLT), pp. 1655–1658, 2016
Trig identities that I used in high school and college
Please support journals that allow authors to post their final papers.
SIAM “Information for Authors”. On their copyright transfer form, they state that the author retains the right to post the final paper on their academic website, as long as the SIAM copyright notice is posted and no fee is charged.
American Physical Society (APS) posting policy. APS publishes the Physical Review series. Summary: authors can post the paper as long as the citation and APS copyright are included, and no fee is charged. The copyright and citation look like: “Authors names, journal title, volume number, page number (or article identifier), year of publication. ”Copyright (year) by the American Physical Society.“”
IEEE Electronic Information Dissemination page. Authors can post a published paper, but need to include the IEEE copyright notice and full citation (see the link for the copyright notice).
The SHERPA/ROMEO website lists the policies of many publishers. The SPARC addendum is an addendum to copyright agreements that authors may request their publisher to sign.