Zobrazeno 1 - 6
of 6
pro vyhledávání: '"Dadi, Leello Tadesse"'
Given a sequence of functions $f_1,\ldots,f_n$ with $f_i:\mathcal{D}\mapsto \mathbb{R}$, finite-sum minimization seeks a point ${x}^\star \in \mathcal{D}$ minimizing $\sum_{j=1}^n f_j(x)/n$. In this work, we propose a key twist into the finite-sum mi
Externí odkaz:
http://arxiv.org/abs/2406.04731
We propose an adaptive variance-reduction method, called AdaSpider, for minimization of $L$-smooth, non-convex functions with a finite-sum structure. In essence, AdaSpider combines an AdaGrad-inspired [Duchi et al., 2011, McMahan & Streeter, 2010], b
Externí odkaz:
http://arxiv.org/abs/2211.01851
Polynomial neural networks (PNNs) have been recently shown to be particularly effective at image generation and face recognition, where high-frequency information is critical. Previous studies have revealed that neural networks demonstrate a $\textit
Externí odkaz:
http://arxiv.org/abs/2202.13473
Autor:
Latorre, Fabian, Krawczuk, Igor, Dadi, Leello Tadesse, Pethick, Thomas Michaelsen, Cevher, Volkan
Adversarial Training using a strong first-order adversary (PGD) is the gold standard for training Deep Neural Networks that are robust to adversarial examples. We show that, contrary to the general understanding of the method, the gradient at an opti
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=od_______185::8ae59200751cd5387c48e72942ae4a51
https://infoscience.epfl.ch/record/300850
https://infoscience.epfl.ch/record/300850
It has been recently observed that neural networks, unlike kernel methods, enjoy a reduced sample complexity when the distribution is isotropic (i.e., when the covariance matrix is the identity). We find that this sensitivity to the data distribution
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=od_______185::8e4913824755f45fb08575c9a6ba9b74
https://infoscience.epfl.ch/record/289659
https://infoscience.epfl.ch/record/289659
We propose an adaptive variance-reduction method, called AdaSpider, for minimization of $L$-smooth, non-convex functions with a finite-sum structure. In essence, AdaSpider combines an AdaGrad-inspired [Duchi et al., 2011, McMahan & Streeter, 2010], b
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::b3b19d6dc77fa1d501bbb3309531beb8
https://infoscience.epfl.ch/record/300277
https://infoscience.epfl.ch/record/300277