Zobrazeno 1 - 10
of 17
pro vyhledávání: '"Ma, Jerry"'
Autor:
Beliveau, Scott, Ma, Jerry
The USPTO disseminates one of the largest publicly accessible repositories of scientific, technical, and commercial data worldwide. USPTO data has historically seen frequent use in fields such as patent analytics, economics, and prosecution & litigat
Externí odkaz:
http://arxiv.org/abs/2207.05239
Publikováno v:
International Conference on Learning Representations (ICLR), 2020
We propose an energy-based model (EBM) of protein conformations that operates at atomic scale. The model is trained solely on crystallized protein data. By contrast, existing approaches for scoring conformations use energy functions that incorporate
Externí odkaz:
http://arxiv.org/abs/2004.13167
Autor:
Ma, Jerry, Yarats, Denis
Adaptive optimization algorithms such as Adam are widely used in deep learning. The stability of such algorithms is often improved with a warmup schedule for the learning rate. Motivated by the difficulty of choosing and tuning warmup schedules, rece
Externí odkaz:
http://arxiv.org/abs/1910.04209
Autor:
Tian, Yuandong, Ma, Jerry, Gong, Qucheng, Sengupta, Shubho, Chen, Zhuoyuan, Pinkerton, James, Zitnick, C. Lawrence
The AlphaGo, AlphaGo Zero, and AlphaZero series of algorithms are remarkable demonstrations of deep reinforcement learning's capabilities, achieving superhuman performance in the complex game of Go with progressively increasing autonomy. However, man
Externí odkaz:
http://arxiv.org/abs/1902.04522
Autor:
Ma, Jerry, Yarats, Denis
Momentum-based acceleration of stochastic gradient descent (SGD) is widely used in deep learning. We propose the quasi-hyperbolic momentum algorithm (QHM) as an extremely simple alteration of momentum SGD, averaging a plain SGD step with a momentum s
Externí odkaz:
http://arxiv.org/abs/1810.06801
Autor:
Ma, Jerry, Yarats, Denis
Publikováno v:
Proceedings of the AAAI Conference on Artificial Intelligence. 35:8828-8836
Adaptive optimization algorithms such as Adam are widely used in deep learning. The stability of such algorithms is often improved with a warmup schedule for the learning rate. Motivated by the difficulty of choosing and tuning warmup schedules, rece
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::9e75a7fe33e24ca331ff96241659f0d6
https://doi.org/10.1609/aaai.v35i10.17069
https://doi.org/10.1609/aaai.v35i10.17069
Akademický článek
Tento výsledek nelze pro nepřihlášené uživatele zobrazit.
K zobrazení výsledku je třeba se přihlásit.
K zobrazení výsledku je třeba se přihlásit.
Publikováno v:
2014 IEEE Symposium on Security & Privacy; 2014, p689-704, 16p
Autor:
Poirier, Maxime, Boudreau, Marcel, Yu-Min Lin, Narayan, Raghuram, Chen, Chengkun, Hong, Xiaoyu, Olson, Rad, Liu, Xu, Gokhale, Milind, Young Ma, Jerry, Eshelman, Mark
Publikováno v:
2015 Optical Fiber Communications Conference & Exhibition (OFC); 2015, p1-3, 3p