no code implementations • 4 Apr 2024 • Ingo Steinwart
In this paper we investigate the conditional distributions of two Banach space valued, jointly Gaussian random variables.
1 code implementation • NeurIPS 2023 • Moritz Haas, David Holzmüller, Ulrike Von Luxburg, Ingo Steinwart
In this paper, we show that the smoothness of the estimators, and not the dimension, is the key: benign overfitting is possible if and only if the estimator's derivatives are large enough.
1 code implementation • 23 Dec 2022 • Marvin Pförtner, Ingo Steinwart, Philipp Hennig, Jonathan Wenger
Crucially, this probabilistic viewpoint allows to (1) quantify the inherent discretization error; (2) propagate uncertainty about the model parameters to the solution; and (3) condition on noisy measurements.
1 code implementation • 23 Jun 2022 • Manuel Nonnenmacher, Lukas Oldenburg, Ingo Steinwart, David Reeb
We therefore devise ExpCLR, a novel contrastive learning approach built on an objective that utilizes expert features to encourage both properties for the learned representation.
2 code implementations • 17 Mar 2022 • David Holzmüller, Viktor Zaverkin, Johannes Kästner, Ingo Steinwart
We provide open-source code that includes efficient implementations of all kernels, kernel transformations, and selection methods, and can be used for reproducing our results.
1 code implementation • NeurIPS 2021 • Manuel Nonnenmacher, Thomas Pfeil, Ingo Steinwart, David Reeb
We validate SOSP-H by comparing it to our second method SOSP-I that uses a well-established Hessian approximation, and to numerous state-of-the-art methods.
1 code implementation • 20 Sep 2021 • Viktor Zaverkin, David Holzmüller, Ingo Steinwart, Johannes Kästner
Artificial neural networks (NNs) are one of the most frequently used machine learning approaches to construct interatomic potentials and enable efficient large-scale atomistic simulations with almost ab initio accuracy.
no code implementations • 4 Nov 2020 • Manuel Nonnenmacher, David Reeb, Ingo Steinwart
The loss surface of an overparameterized neural network (NN) possesses many global minima of zero training error.
1 code implementation • 12 Feb 2020 • David Holzmüller, Ingo Steinwart
We prove that two-layer (Leaky)ReLU networks initialized by e. g. the widely used method proposed by He et al. (2015) and trained using gradient descent on a least-squares loss are not universally consistent.
no code implementations • 8 Feb 2020 • Ingo Steinwart
Given an uncountable, compact metric space, we show that there exists no reproducing kernel Hilbert space that contains the space of all continuous functions on this compact space.
no code implementations • 25 Sep 2019 • Manuel Nonnenmacher, David Reeb, Ingo Steinwart
The recently developed link between strongly overparametrized neural networks (NNs) and kernel methods has opened a new way to understand puzzling features of NNs, such as their convergence and generalization behaviors.
no code implementations • 27 May 2019 • Hanyuan Hang, Xiaoyu Liu, Ingo Steinwart
We propose an algorithm named best-scored random forest for binary classification problems.
no code implementations • 25 May 2019 • Nicole Mücke, Ingo Steinwart
Moreover, we show that the same phenomenon occurs for DNNs with zero training error and sufficiently large architectures.
no code implementations • 27 Mar 2019 • Ingo Steinwart
Initializing the weights and the biases is a key part of the training process of a neural network.
no code implementations • 4 Oct 2018 • Hanyuan Hang, Ingo Steinwart
This paper investigates the nonparametric regression problem using SVMs with anisotropic Gaussian RBF kernels.
no code implementations • 14 Dec 2017 • Ingo Steinwart, Johanna F. Ziegel
Strictly proper kernel scores are well-known tool in probabilistic forecasting, while characteristic kernels have been extensively investigated in the machine learning literature.
no code implementations • 17 Aug 2017 • Ingo Steinwart, Bharath K. Sriperumbudur, Philipp Thomann
We derive and analyze a generic, recursive algorithm for estimating all splits in a finite cluster tree as well as the corresponding clusters.
no code implementations • 24 Feb 2017 • Muhammad Farooq, Ingo Steinwart
Conditional expectiles are becoming an increasingly important tool in finance as well as in other areas of applications.
no code implementations • 23 Feb 2017 • Simon Fischer, Ingo Steinwart
In this paper we extend these rates to norms stronger than the $L_2$-norm without requiring the regression function to be contained in the hypothesis space.
1 code implementation • 22 Feb 2017 • Ingo Steinwart, Philipp Thomann
liquidSVM is a package written in C++ that provides SVM-type solvers for various classification and regression tasks.
no code implementations • 2 Dec 2016 • Ingo Steinwart, Philipp Thomann, Nico Schmid
We investigate iterated compositions of weighted sums of Gaussian kernels and provide an interpretation of the construction that shows some similarities with the architectures of deep neural networks.
no code implementations • 1 Dec 2016 • Philipp Thomann, Ingrid Blaschzyk, Mona Meister, Ingo Steinwart
Our contributions are two fold: On the theoretical side we establish an oracle inequality for the overall learning method using the hinge loss, and show that the resulting rates match those known for SVMs solving the complete optimization problem with Gaussian kernels.
no code implementations • 13 Jul 2016 • Hanyuan Hang, Ingo Steinwart, Yunlong Feng, Johan A. K. Suykens
We study the density estimation problem with observations generated by certain dynamical systems that admit a unique underlying invariant Lebesgue density.
no code implementations • 10 May 2016 • Hanyuan Hang, Yunlong Feng, Ingo Steinwart, Johan A. K. Suykens
We show that when the stochastic processes satisfy a generalized Bernstein-type inequality, a unified treatment on analyzing the learning schemes with various mixing processes can be conducted and a sharp oracle inequality for generic regularized empirical risk minimization schemes can be established.
no code implementations • 21 Aug 2015 • Ingo Steinwart
We characterize when the level sets of a continuous quasi-monotone functional defined on a suitable convex subset of a normed space can be uniquely represented by a family of bounded continuous functionals.
no code implementations • 15 Aug 2015 • Philipp Thomann, Ingo Steinwart, Nico Schmid
We propose some axioms for hierarchical clustering of probability measures and investigate their ramifications.
no code implementations • 23 Jul 2015 • Mona Eberts, Ingo Steinwart
In addition, their motivation was always based on computational requirements.
no code implementations • 14 Jul 2015 • Muhammad Farooq, Ingo Steinwart
Expectile regression is a nice tool for investigating conditional distributions beyond the conditional mean.
no code implementations • 30 Sep 2014 • Ingo Steinwart
The clusters of a distribution are often defined by the connected components of a density level set.
no code implementations • NeurIPS 2011 • Mona Eberts, Ingo Steinwart
We prove a new oracle inequality for support vector machines with Gaussian RBF kernels solving the regularized least squares regression problem.
no code implementations • NeurIPS 2010 • Andreas Christmann, Ingo Steinwart
We apply this technique for the following special cases: universal kernels on the set of probability measures, universal kernels based on Fourier transforms, and universal kernels for signal processing.
no code implementations • NeurIPS 2009 • Ingo Steinwart, Andreas Christmann
We prove an oracle inequality for generic regularized empirical risk minimization algorithms learning from $\a$-mixing processes.
no code implementations • NeurIPS 2008 • Ingo Steinwart, Andreas Christmann
In this paper lower and upper bounds for the number of support vectors are derived for support vector machines (SVMs) based on the epsilon-insensitive loss function.