Zobrazeno 1 - 10
of 61
pro vyhledávání: '"Suvrit Sra"'
Autor:
Ramkumar Hariharan, Johnna Sundberg, Giacomo Gallino, Ashley Schmidt, Drew Arenth, Suvrit Sra, Benjamin Fels
Publikováno v:
Frontiers in Artificial Intelligence, Vol 3 (2020)
Providing accurate utilization forecasts is key to maintaining optimal vaccine stocks in any health facility. Current approaches to vaccine utilization forecasting are based on often outdated population census data, and rely on weak, low-dimensional
Externí odkaz:
https://doaj.org/article/457fc8c3e7cf4277989edb6b5a7cfe73
Autor:
Suvrit Sra, Melanie Weber
Publikováno v:
IMA Journal of Numerical Analysis. 42:3241-3271
We study stochastic projection-free methods for constrained optimization of smooth functions on Riemannian manifolds, i.e., with additional constraints beyond the parameter domain being a manifold. Specifically, we introduce stochastic Riemannian Fra
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::90ad33c5fff667f7009008fa481a143f
https://doi.org/10.1093/imanum/drab066
https://doi.org/10.1093/imanum/drab066
Autor:
Melanie Weber, Suvrit Sra
We study projection-free methods for constrained Riemannian optimization. In particular, we propose the Riemannian Frank-Wolfe (RFW) method. We analyze non-asymptotic convergence rates of RFW to an optimum for (geodesically) convex problems, and to a
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::fa50696011e5109e9ed6457b1658c691
https://doi.org/10.1007/s10107-022-01840-5
https://doi.org/10.1007/s10107-022-01840-5
Publikováno v:
Adv Neural Inf Process Syst
The generalization of representations learned via contrastive learning depends crucially on what features of the data are extracted. However, we observe that the contrastive loss does not always sufficiently guide which features are extracted, a beha
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::4fd5d3cac1523f9c1cb3e8834d3a9d27
Autor:
Suvrit Sra, Reshad Hosseini
Publikováno v:
Handbook of Variational Methods for Nonlinear Geometric Data ISBN: 9783030313500
Stochastic and finite-sum optimization problems are central to machine learning. Numerous specializations of these problems involve nonlinear constraints where the parameters of interest lie on a manifold. Consequently, stochastic manifold optimizati
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::acfc80335f58274c04c2228b2e0857de
https://doi.org/10.1007/978-3-030-31351-7_19
https://doi.org/10.1007/978-3-030-31351-7_19
Autor:
Suvrit Sra
Publikováno v:
arXiv
We prove some "power" generalizations of Marcus-Lopes-style (including McLeod and Bullen) concavity inequalities for elementary symmetric polynomials, and convexity inequalities (of McLeod and Baston) for complete homogeneous symmetric polynomials. F
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::aba05d7ca53e0339e6984e96d04548f4
https://hdl.handle.net/1721.1/136374.2
https://hdl.handle.net/1721.1/136374.2
Autor:
Suvrit Sra
Publikováno v:
Proceedings of the American Mathematical Society. 147:481-486
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::f2635db9f1328a698db6afeec5916465
https://doi.org/10.1090/proc/14023
https://doi.org/10.1090/proc/14023
Publikováno v:
Linear Algebra and its Applications. 528:124-146
We prove the sum of squared logarithms inequality (SSLI) which states that for nonnegative vectors x , y ∈ R n whose elementary symmetric polynomials satisfy e k ( x ) ≤ e k ( y ) (for 1 ≤ k n ) and e n ( x ) = e n ( y ) , the inequality ∑ i
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::9df48d51f0b895d125550afb5d025381
https://doi.org/10.1016/j.laa.2016.06.026
https://doi.org/10.1016/j.laa.2016.06.026
Autor:
Suvrit Sra
We study metric properties of symmetric divergences on Hermitian positive definite matrices. In particular, we prove that the square root of these divergences is a distance metric. As a corollary we obtain a proof of the metric property for Quantum J
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::14fe3760015f1e10cab89d48fbc85840
http://arxiv.org/abs/1911.02643
http://arxiv.org/abs/1911.02643