Zobrazeno 1 - 10
of 895
pro vyhledávání: '"Cohen, Jeremy A."'
Optimization in deep learning remains poorly understood, even in the simple setting of deterministic (i.e. full-batch) training. A key difficulty is that much of an optimizer's behavior is implicitly determined by complex oscillatory dynamics, referr
Externí odkaz:
http://arxiv.org/abs/2410.24206
Publikováno v:
Diversity and Distributions, 2024 Aug 01. 30(8), 1-13.
Externí odkaz:
https://www.jstor.org/stable/48780051
Autor:
Cohen, Jeremy E., Leplat, Valentin
Regularized nonnegative low-rank approximations such as sparse Nonnegative Matrix Factorization or sparse Nonnegative Tucker Decomposition are an important branch of dimensionality reduction models with enhanced interpretability. However, from a prac
Externí odkaz:
http://arxiv.org/abs/2403.18517
Autor:
Goth, Florian, Alves, Renato, Braun, Matthias, Castro, Leyla Jael, Chourdakis, Gerasimos, Christ, Simon, Cohen, Jeremy, Druskat, Stephan, Erxleben, Fredo, Grad, Jean-Noël, Hagdorn, Magnus, Hodges, Toby, Juckeland, Guido, Kempf, Dominic, Lamprecht, Anna-Lena, Linxweiler, Jan, Löffler, Frank, Martone, Michele, Schwarzmeier, Moritz, Seibold, Heidi, Thiele, Jan Philipp, von Waldow, Harald, Wittke, Samantha
The term Research Software Engineer, or RSE, emerged a little over 10 years ago as a way to represent individuals working in the research community but focusing on software development. The term has been widely adopted and there are a number of high-
Externí odkaz:
http://arxiv.org/abs/2311.11457
Nonnegative Matrix Factorization is an important tool in unsupervised machine learning to decompose a data matrix into a product of parts that are often interpretable. Many algorithms have been proposed during the last three decades. A well-known met
Externí odkaz:
http://arxiv.org/abs/2303.17992
Structural segmentation of music refers to the task of finding a symbolic representation of the organisation of a song, reducing the musical flow to a partition of non-overlapping segments. Under this definition, the musical structure may not be uniq
Externí odkaz:
http://arxiv.org/abs/2212.11054
Music Structure Analysis (MSA) consists of representing a song in sections (such as ``chorus'', ``verse'', ``solo'' etc), and can be seen as the retrieval of a simplified organization of the song. This work presents a new algorithm, called Convolutiv
Externí odkaz:
http://arxiv.org/abs/2210.15356
Autor:
Kobayashi, Rika, Jaffa, Sarah, Dong, Jiachen, Amos, Roger D., Cohen, Jeremy, Kerrison, Emily F.
A virtual hands-on computer laboratory has been designed within the Gather online meeting platform. Gather's features such as spatial audio, private spaces and interactable objects offer scope for great improvements over currently used platforms, esp
Externí odkaz:
http://arxiv.org/abs/2209.01927
Autor:
GOLDSTEIN, ADAM, EATON, CHARLIE, VILLALOBOS, AMBER, CHAKRABARTI, PARIJAT, COHEN, JEREMY, DONNELLY, KATIE
Publikováno v:
RSF: The Russell Sage Foundation Journal of the Social Sciences, 2023 Sep 01. 9(4), 86-112.
Externí odkaz:
https://www.jstor.org/stable/48738305
Autor:
Cohen, Jeremy M., Ghorbani, Behrooz, Krishnan, Shankar, Agarwal, Naman, Medapati, Sourabh, Badura, Michal, Suo, Daniel, Cardoze, David, Nado, Zachary, Dahl, George E., Gilmer, Justin
Very little is known about the training dynamics of adaptive gradient methods like Adam in deep learning. In this paper, we shed light on the behavior of these algorithms in the full-batch and sufficiently large batch settings. Specifically, we empir
Externí odkaz:
http://arxiv.org/abs/2207.14484