New deep neural network design bypasses curse of dimensionality for function approximation

Researchers have developed symmetric deep neural networks capable of approximating symmetric Korobov functions. The study demonstrates that the convergence rate and prefactor in these approximations scale polynomially with the ambient dimension, a significant improvement over previous methods that were subject to the curse of dimensionality. This work also establishes a generalization-error rate for learning symmetric Korobov functions that avoids this dimensional dependency. AI

IMPACT This research offers a theoretical advancement in how deep neural networks can approximate complex functions, potentially leading to more efficient and scalable models for certain applications.

RANK_REASON Academic paper detailing a new theoretical approach to deep neural network approximation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• alphaXiv
• arXiv
• CatalyzeX Code Finder for Papers
• Connected Papers
• CORE Recommender
• DagsHub
• Deep Neural Networks
• Gotit.pub
• Hugging Face
• IArxiv Recommender
• Influence Flower
• Korobov Functions
• Litmaps
• ScienceCast
• scite Smart Citations
• Yahong Yang

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →