Zobrazeno 1 - 9
of 9
pro vyhledávání: '"Preetum Nakkiran"'
We study the fundamental question of how to define and measure the distance from calibration for probabilistic predictors. While the notion of perfect calibration is well-understood, there is no consensus on how to quantify the distance from perfect
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::9d4fe4133219651727418497477ed25d
http://arxiv.org/abs/2211.16886
http://arxiv.org/abs/2211.16886
Autor:
Kasper Green Larsen, Jaroslaw Blasiok, Michael Mitzenmacher, Preetum Nakkiran, Ben Lawson, Charalampos E. Tsourakakis, Vasileios Nakos
Publikováno v:
Internet Mathematics.
Social networks involve both positive and negative relationships, which can be captured in signed graphs. The {\em edge sign prediction problem} aims to predict whether an interaction between a pair of nodes will be positive or negative. We provide t
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::72ce64f16ed8038793f857cf8e38af30
https://doi.org/10.24166/im.05.2019
https://doi.org/10.24166/im.05.2019
Autor:
Shibani Santurkar, Reiichiro Nakano, Logan Engstrom, Justin Gilmer, Brandon Tran, Preetum Nakkiran, Dimitris Tsipras, Andrew Ilyas, Dan Hendrycks, Eric Wallace, Aleksander Madry, Gabriel Goh
Publikováno v:
Distill. 4
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::16454985094ddea19f819455c0a0baf8
https://doi.org/10.23915/distill.00019
https://doi.org/10.23915/distill.00019
Autor:
Preetum Nakkiran
Publikováno v:
Distill. 4
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::1890a792e644b0d1c5fdc9bc9b745628
https://doi.org/10.23915/distill.00019.5
https://doi.org/10.23915/distill.00019.5
Publikováno v:
STOC
Arikan's exciting discovery of polar codes has provided an altogether new way to efficiently achieve Shannon capacity. Given a (constant-sized) invertible matrix $M$, a family of polar codes can be associated with this matrix and its ability to appro
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::abc8b85f0fb6e6a57233a620b984f94e
https://doi.org/10.1145/3188745.3188816
https://doi.org/10.1145/3188745.3188816
Publikováno v:
INTERSPEECH
We present a general approach to reduce the size of feedforward deep neural networks (DNNs). We propose a rankconstrained topology, which factors the weights in the input layer of the DNN in terms of a low-rank representation: unlike previous work, o
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::65f7c0aab74e333aa9a0b4ac55d1e535
https://doi.org/10.21437/interspeech.2015-351
https://doi.org/10.21437/interspeech.2015-351
Publikováno v:
ISIT
Erasure codes are being increasingly used in distributed-storage systems in place of data-replication, since they provide the same level of reliability with much lower storage overhead. We consider the problem of constructing explicit erasure codes f
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::a702834bac88ba9b2a99008021d78153
http://arxiv.org/abs/1509.01858
http://arxiv.org/abs/1509.01858
Publikováno v:
ICASSP
We explore techniques to improve the robustness of small-footprint keyword spotting models based on deep neural networks (DNNs) in the presence of background noise and in far-field conditions. We find that system performance can be improved significa
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::438111b8d333240fdbc1ee50eb4a6d0f
https://doi.org/10.1109/icassp.2015.7178863
https://doi.org/10.1109/icassp.2015.7178863
Publikováno v:
GLOBECOM
In distributed storage systems, storage nodes intermittently go offline for numerous reasons. On coming back online, nodes need to update their contents to reflect any modifications to the data in the interim. In this paper, we consider a setting whe
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::47ecbbfc766571b7fe476f8b141e9ece
http://arxiv.org/abs/1409.1666
http://arxiv.org/abs/1409.1666