paper-with-me

홈 › Papers

Adaptive Approximation and Generalization of Deep Neural Network with Intrinsic Dimensionality

2019-07-04 · Ryumei Nakada, Masaaki Imaizumi

In this study, we prove that an intrinsic low dimensionality of covariates is the main factor that determines the performance of deep neural networks (DNNs). DNNs generally provide outstanding empirical performance. Hence, numerous studies have actively investigated the theoretical properties of DNNs to understand their underlying mechanisms. In particular, the behavior of DNNs in terms of high-dimensional data is one of the most critical questions. However, this issue has not been sufficiently investigated from the aspect of covariates, although high-dimensional data have practically low intrinsic dimensionality. In this study, we derive bounds for an approximation error and a generalization error regarding DNNs with intrinsically low dimensional covariates. We apply the notion of the Minkowski dimension and develop a novel proof technique. Consequently, we show that convergence rates of the errors by DNNs do not depend on the nominal high dimensionality of data, but on its lower intrinsic dimension. We further prove that the rate is optimal in the minimax sense. We identify an advantage of DNNs by showing that DNNs can handle a broader class of intrinsic low dimensional data than other adaptive estimators. Finally, we conduct a numerical simulation to validate the theoretical results.

📄 PDF Abstract BibTeX arXiv:1907.02177

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Adaptive Metric Dimensionality Reduction

2013-02-12 · Lee-Ad Gottlieb, Aryeh Kontorovich, Robert Krauthgamer

We study adaptive data-dependent dimensionality reduction in the context of supervised learning in general metric spaces. Our main statistical contribution is a generalization bound for Lipschitz functions in metric spac…

Dimensionality ReductionGeneralization Bounds

Deep learning is adaptive to intrinsic dimensionality of model smoothness in anisotropic Besov space

2019-10-28 · NeurIPS 2021 12 · Taiji Suzuki, Atsushi Nitanda

Deep learning has exhibited superior performance for various tasks, especially for high-dimensional datasets, such as images. To understand this property, we investigate the approximation and estimation ability of deep l…

Deep Learning

Effective Reasoning Chains Reduce Intrinsic Dimensionality

2026-02-09 · Archiki Prasad, Mandar Joshi, Kenton Lee, Mohit Bansal 외 arxiv

Chain-of-thought (CoT) reasoning and its variants have substantially improved the performance of language models on complex reasoning tasks, yet the precise mechanisms by which different strategies facilitate generalizat…

From Low Intrinsic Dimensionality to Non-Vacuous Generalization Bounds in Deep Multi-Task Learning

2025-01-31 · Hossein Zakerinia, Dorsa Ghobadi, Christoph H. Lampert

Deep learning methods are known to generalize well from training to future data, even in an overparametrized regime, where they could easily overfit. One explanation for this phenomenon is that even when their *ambient d…

Generalization BoundsMulti-Task Learning

A deep network construction that adapts to intrinsic dimensionality beyond the domain

2020-08-06 · Alexander Cloninger, Timo Klock

We study the approximation of two-layer compositions $f(x) = g(\phi(x))$ via deep networks with ReLU activation, where $\phi$ is a geometrically intuitive, dimensionality reducing feature map. We focus on two intuitive a…