English
Related papers

Related papers: The Curse of Instability

200 papers

This paper demonstrates that when a shallow neural network with a Lipschitz continuous activation function is trained using either empirical or population risk to approximate a target function that is $r$ times continuously differentiable…

Machine Learning · Computer Science 2026-03-06 Sanghoon Na , Haizhao Yang

We prove the curse of dimensionality in the worst case setting for multivariate numerical integration for various classes of smooth functions. We prove the results when the domains are isotropic convex bodies with small diameter satisfying…

Numerical Analysis · Mathematics 2019-08-15 Aicke Hinrichs , Joscha Prochno , Mario Ullrich

The failure of the Euclidean norm to reliably distinguish between nearby and distant points in high dimensional space is well-known. This phenomenon of distance concentration manifests in a variety of data distributions, with iid or…

Machine Learning · Computer Science 2022-01-17 Brian A. Powell

Today's best language models still struggle with hallucinations: factually incorrect generations, which impede their ability to reliably retrieve information seen during training. The reversal curse, where models cannot recall information…

Machine Learning · Computer Science 2024-06-11 Ouail Kitouni , Niklas Nolte , Diane Bouchacourt , Adina Williams , Mike Rabbat , Mark Ibrahim

As is known, factor analysis is a popular method to reduce dimension for high-dimensional data. For matrix data, the dimension reduction can be more effectively achieved through both row and column directions. In this paper, we introduce a…

Methodology · Statistics 2019-04-17 Xialu Liu , Elynn Chen

A regression problem with dependent data is considered. Regularity assumptions on the dependency of the data are introduced, and it is shown that under suitable structural assumptions on the regression function a deep recurrent neural…

Machine Learning · Statistics 2020-11-03 Michael Kohler , Adam Krzyzak

We present an algorithm (CoDeFi) which overcomes the curse of dimensionality (CoD) in scientific computations and, especially, in mathematical finance (Fi). Our method applies a broad class of partial differential equations such as…

Numerical Analysis · Mathematics 2016-10-07 Philippe G. LeFloch , Jean-Marc Mercier

It has long been noticed that high dimension data exhibits strange patterns. This has been variously interpreted as either a "blessing" or a "curse", causing uncomfortable inconsistencies in the literature. We propose that these patterns…

Computer Vision and Pattern Recognition · Computer Science 2020-03-18 Wen-Yan Lin

We offer a theoretical validation of the curse of dimensionality in the pivot-based indexing of datasets for similarity search, by proving, in the framework of statistical learning, that in high dimensions no pivot-based indexing scheme can…

Data Structures and Algorithms · Computer Science 2016-11-17 Ilya Volnyansky , Vladimir Pestov

Most approximations for stochastic differential equations with high-dimensional, non-Gaussian inputs suffer from a rapid (e.g., exponential) increase of computational cost, an issue known as the curse of dimensionality. In astrodynamics,…

Applications · Statistics 2017-08-30 Marc Balducci , Brandon Jones , Alireza Doostan

This chapter does not deal with specific tools and techniques for managing complex systems, but proposes some basic concepts that help us to think and speak about complexity. We review classical thinking and its intrinsic drawbacks when…

Adaptation and Self-Organizing Systems · Physics 2007-05-23 Carlos Gershenson , Francis Heylighen

Recent studies highlight economic complexity's role in mitigating fiscal crises, often measured via an economy's trade structure. Trade, however, is just one facet of an economy's structure and omits critical innovative activities like…

General Economics · Economics 2024-11-05 Goran Hristovski , Gjorgji Gockov , Viktor Stojkoski

In this work we study the validity of the so-called curse of dimensionality for indexing of databases for similarity search. We perform an asymptotic analysis, with a test model based on a sequence of metric spaces $(\Omega_d)$ from which…

Data Structures and Algorithms · Computer Science 2009-05-14 Ilya Volnyansky

We show that the minimal discrepancy of a point set in the $d$-dimensional unit cube with respect to the BMO seminorm suffers from the curse of dimensionality.

Number Theory · Mathematics 2023-02-02 Friedrich Pillichshammer

Usually when applying the mimetic model to the early universe, higher derivative terms are needed to promote the mimetic field to be dynamical. However such models suffer from the ghost and/or the gradient instabilities and simple…

General Relativity and Quantum Cosmology · Physics 2017-09-06 Yunlong Zheng , Liuyuan Shen , Yicen Mou , Mingzhe Li

Deep learning algorithms are responsible for a technological revolution in a variety of tasks including image recognition or Go playing. Yet, why they work is not understood. Ultimately, they manage to classify data lying in high dimension…

Machine Learning · Computer Science 2021-01-01 Mario Geiger , Leonardo Petrini , Matthieu Wyart

We find that the performance of state-of-the-art models on Natural Language Inference (NLI) and Reading Comprehension (RC) analysis/stress sets can be highly unstable. This raises three questions: (1) How will the instability affect the…

Computation and Language · Computer Science 2020-11-17 Xiang Zhou , Yixin Nie , Hao Tan , Mohit Bansal

In gravitational theories with extra dimensions, it is argued that the existence of a positive vacuum energy generically implies catastrophic instability of our four-dimensional world. The most generic instability is a decompactification…

High Energy Physics - Theory · Physics 2009-11-10 Steven B. Giddings

Difficulties in replication and reproducibility of empirical evidences in machine learning research have become a prominent topic in recent years. Ensuring that machine learning research results are sound and reliable requires…

Machine Learning · Computer Science 2024-03-20 Tobias Hille , Maximilian Stubbemann , Tom Hanika

The dynamics of a nonequilibrium system can become complex because the system has many components (e.g., a human brain), because the system is strongly driven from equilibrium (e.g., large Reynolds-number flows), or because the system…

chao-dyn · Physics 2008-02-03 Henry S. Greenside