Zobrazeno 1 - 10
of 124
pro vyhledávání: '"Michael W, Mahoney"'
Publikováno v:
Communications Physics, Vol 6, Iss 1, Pp 1-14 (2023)
Abstract Dynamical systems that evolve continuously over time are ubiquitous throughout science and engineering. Machine learning (ML) provides data-driven approaches to model and predict the dynamics of such systems. A core issue with this approach
Externí odkaz:
https://doaj.org/article/776f47981bd7428aab296c6fe960eb7f
Publikováno v:
Nature Communications, Vol 12, Iss 1, Pp 1-13 (2021)
In many machine learning applications, one uses pre-trained neural networks, having limited access to training and test data. Martin et al. show how to predict trends in the quality of such neural networks without access to this information, relevant
Externí odkaz:
https://doaj.org/article/bdc9afbd811d47888c4645cf78e0b595
Publikováno v:
EURO Journal on Computational Optimization, Vol 10, Iss , Pp 100035- (2022)
We consider a variant of inexact Newton Method [20,40], called Newton-MR, in which the least-squares sub-problems are solved approximately using Minimum Residual method [79]. By construction, Newton-MR can be readily applied for unconstrained optimiz
Externí odkaz:
https://doaj.org/article/e1dc657164394223a91e4ff3c91e58f7
Publikováno v:
SIAM Review. 65:59-143
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::102ed86e53048f87fa08601ac87cae34
https://doi.org/10.1137/20m1333055
https://doi.org/10.1137/20m1333055
Autor:
Surajit Das, Laura Grigori, Kimon Fountoulakis, James Demmel, Michael W. Mahoney, Shenghao Yang
Publikováno v:
SIAM Journal on Scientific Computing. 43:C154-C176
We are interested in parallelizing the Least Angle Regression (LARS) algorithm for fitting linear regression models to high-dimensional data. We consider two parallel and communication avoiding versions of the basic LARS algorithm. The two algorithms
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::2b03cb639066fff47a36c10267e32641
https://doi.org/10.1137/19m1305720
https://doi.org/10.1137/19m1305720
Publikováno v:
INFORMS Journal on Optimization. 3:154-182
The paper aims to extend the theory and application of nonconvex Newton-type methods, namely trust region and cubic regularization, to the settings in which, in addition to the solution of subproblems, the gradient and the Hessian of the objective fu
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::7f00125da32fcb8732b389535cbec5ee
https://doi.org/10.1287/ijoo.2019.0043
https://doi.org/10.1287/ijoo.2019.0043
Autor:
Zhewei Yao, Linjian Ma, Kurt Keutzer, Jiayu Ye, Zhen Dong, Amir Gholami, Michael W. Mahoney, Sheng Shen
Publikováno v:
AAAI
Transformer based architectures have become de-facto models used for a range of Natural Language Processing tasks. In particular, the BERT based models achieved significant accuracy gain for GLUE tasks, CoNLL-03 and SQuAD. However, BERT based models
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::6c1518c84da710c750919ae90c3d6564
https://doi.org/10.1609/aaai.v34i05.6409
https://doi.org/10.1609/aaai.v34i05.6409
Publikováno v:
Low-Power Computer Vision ISBN: 9781003162810
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::ad38f083f4fbea899c4dc859b9a043c8
https://doi.org/10.1201/9781003162810-13
https://doi.org/10.1201/9781003162810-13
We consider minimizing a smooth and strongly convex objective function using a stochastic Newton method. At each iteration, the algorithm is given an oracle access to a stochastic estimate of the Hessian matrix. The oracle model includes popular algo
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::e321fadf642ddce9f796190e0ef4ec8c
Autor:
Zhengming Zhang, Yaoqing Yang, Zhewei Yao, Yujun Yan, Joseph E. Gonzalez, Kannan Ramchandran, Michael W. Mahoney
Publikováno v:
2021 IEEE International Conference on Big Data (Big Data).
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::0a58687cc80f4f166ca955d6f626d399
https://doi.org/10.1109/bigdata52589.2021.9671693
https://doi.org/10.1109/bigdata52589.2021.9671693