Zobrazeno 1 - 10
of 855
pro vyhledávání: '"van de Geer A"'
Autor:
van de Geer, Sara
We derive entropy bounds for the absolute convex hull of vectors $X= (x_1 , \ldots , x_p)\in \mathbb{R}^{n \times p} $ in $\mathbb{R}^n$ and apply this to the case where $X$ is the $d$-fold tensor matrix $$X = \underbrace{\Psi \otimes \cdots \otimes
Externí odkaz:
http://arxiv.org/abs/2402.08388
The logistic regression estimator is known to inflate the magnitude of its coefficients if the sample size $n$ is small, the dimension $p$ is (moderately) large or the signal-to-noise ratio $1/\sigma$ is large (probabilities of observing a label are
Externí odkaz:
http://arxiv.org/abs/2305.15991
Autor:
Casper H. van de Geer, Annette C. Broderick, Matt I.D. Carter, Athuman Abdallah Irei, Fikiri Kea Kiponda, Joseph Kiptum, Joe Ngunu Wandiga, Mohamed Omar, Nicola Parazzi, Hannah Sawyer-Kerr, Sam B. Weber, Ricardo Zanre, Brendan J. Godley
Publikováno v:
Oryx, Vol 58, Pp 310-322 (2024)
For the Western Indian Ocean region, there is a significant knowledge gap regarding marine turtle nesting on the continental coast of East Africa. Here we present results from a long-term (2000–2020) community-based monitoring programme in and arou
Externí odkaz:
https://doaj.org/article/e9bdeed82ca04bf58f31c4e2d7640780
This paper studies binary classification in robust one-bit compressed sensing with adversarial errors. It is assumed that the model is overparameterized and that the parameter of interest is effectively sparse. AdaBoost is considered, and, through it
Externí odkaz:
http://arxiv.org/abs/2105.02083
Autor:
Ortelli, Francesco, van de Geer, Sara
We extend the notion of trend filtering to tensors by considering the $k^{\rm th}$-order Vitali variation, a discretized version of the integral of the absolute value of the $k^{\rm th}$-order total derivative. We prove adaptive $\ell^0$-rates and no
Externí odkaz:
http://arxiv.org/abs/2101.10692
Autor:
K. de Joode, W. S. van de Geer, G. J. L. H. van Leenders, P. Hamberg, H. M. Westgeest, A. Beeker, S. F. Oosting, J. M. van Rooijen, L. V. Beerepoot, M. Labots, R. H. J. Mathijssen, M. P. Lolkema, E. Cuppen, S. Sleijfer, H. J. G. van de Werken, A. A. M. van der Veldt
Publikováno v:
Scientific Reports, Vol 13, Iss 1, Pp 1-15 (2023)
Abstract Differences in the clinical course and treatment responses in individual patients with advanced renal cell carcinoma (RCC) can largely be explained by the different genomics of this disease. To improve the personalized treatment strategy and
Externí odkaz:
https://doaj.org/article/0bbbda1544a74e2ab59f748e487111a1
This article develops a general theory for minimum norm interpolating estimators and regularized empirical risk minimizers (RERM) in linear models in the presence of additive, potentially adversarial, errors. In particular, no conditions on the error
Externí odkaz:
http://arxiv.org/abs/2012.00807
Autor:
Hinz, Peter, van de Geer, Sara
Feed-forward ReLU neural networks partition their input domain into finitely many "affine regions" of constant neuron activation pattern and affine behaviour. We analyze their mathematical structure and provide algorithmic primitives for an efficient
Externí odkaz:
http://arxiv.org/abs/2011.14895
Autor:
van de Geer, Sara
We study logistic regression with total variation penalty on the canonical parameter and show that the resulting estimator satisfies a sharp oracle inequality: the excess risk of the estimator is adaptive to the number of jumps of the underlying sign
Externí odkaz:
http://arxiv.org/abs/2003.02678
Autor:
Ortelli, Francesco, van de Geer, Sara
Publikováno v:
Journal of Machine Learning Research, 21(247), 2020
We study the theoretical properties of image denoising via total variation penalized least-squares. We define the total vatiation in terms of the two-dimensional total discrete derivative of the image and show that it gives rise to denoised images th
Externí odkaz:
http://arxiv.org/abs/1911.07231