Zobrazeno 1 - 10
of 50
pro vyhledávání: '"Vandermeulen, Robert A."'
We show that deep neural networks achieve dimension-independent rates of convergence for learning structured densities such as those arising in image, audio, video, and text applications. More precisely, we demonstrate that neural networks with a sim
Externí odkaz:
http://arxiv.org/abs/2411.15095
We consider the problem of estimating a structured multivariate density, subject to Markov conditions implied by an undirected graph. In the worst case, without Markovian assumptions, this problem suffers from the curse of dimensionality. Our main re
Externí odkaz:
http://arxiv.org/abs/2410.07685
Autor:
Muttenthaler, Lukas, Vandermeulen, Robert A., Zhang, Qiuyi, Unterthiner, Thomas, Müller, Klaus-Robert
Model overconfidence and poor calibration are common in machine learning and difficult to account for when applying standard empirical risk minimization. In this work, we propose a novel method to alleviate these problems that we call odd-$k$-out lea
Externí odkaz:
http://arxiv.org/abs/2307.02245
Autor:
Muttenthaler, Lukas, Linhardt, Lorenz, Dippel, Jonas, Vandermeulen, Robert A., Hermann, Katherine, Lampinen, Andrew K., Kornblith, Simon
Deep neural networks have reached human-level performance on many computer vision tasks. However, the objectives used to train these networks enforce only that similar images are embedded at similar locations in the representation space, and do not d
Externí odkaz:
http://arxiv.org/abs/2306.04507
Autor:
Vandermeulen, Robert A.
Recent works have demonstrated that the convergence rate of a nonparametric density estimator can be greatly improved by using a low-rank estimator when the target density is a convex combination of separable probability densities with Lipschitz cont
Externí odkaz:
http://arxiv.org/abs/2302.04292
Autor:
Muttenthaler, Lukas, Dippel, Jonas, Linhardt, Lorenz, Vandermeulen, Robert A., Kornblith, Simon
Today's computer vision models achieve human or near-human level performance across a wide variety of vision tasks. However, their architectures, data, and learning algorithms differ in numerous ways from those that give rise to human vision. In this
Externí odkaz:
http://arxiv.org/abs/2211.01201
Recent work has shown that finite mixture models with $m$ components are identifiable, while making no assumptions on the mixture components, so long as one has access to groups of samples of size $2m-1$ which are known to come from the same mixture
Externí odkaz:
http://arxiv.org/abs/2207.11164
Autor:
Liznerski, Philipp, Ruff, Lukas, Vandermeulen, Robert A., Franks, Billy Joe, Müller, Klaus-Robert, Kloft, Marius
Due to the intractability of characterizing everything that looks unlike the normal data, anomaly detection (AD) is traditionally treated as an unsupervised problem utilizing only normal samples. However, it has recently been found that unsupervised
Externí odkaz:
http://arxiv.org/abs/2205.11474
Autor:
Muttenthaler, Lukas, Zheng, Charles Y., McClure, Patrick, Vandermeulen, Robert A., Hebart, Martin N., Pereira, Francisco
A central goal in the cognitive sciences is the development of numerical models for mental representations of object concepts. This paper introduces Variational Interpretable Concept Embeddings (VICE), an approximate Bayesian method for embedding obj
Externí odkaz:
http://arxiv.org/abs/2205.00756
The construction and theoretical analysis of the most popular universally consistent nonparametric density estimators hinge on one functional property: smoothness. In this paper we investigate the theoretical implications of incorporating a multi-vie
Externí odkaz:
http://arxiv.org/abs/2204.00930