New Paper: Locally Adaptive Neural Networks Hit Pointwise Minimax-Optimal Rates
udmrzn · x · 2026-11-18
An arXiv paper proposes training a local neural network on the k nearest neighbors of each query point instead of fitting one global model, addressing regression functions with spatially varying smoothness.
Key contributions:
- For locally α-Hölder regression functions on a d-dimensional manifold, the pointwise risk attains rate n^{-2α/(2α+d)} up to log factors, shown to be pointwise minimax optimal via a matching Le Cam lower bound.
- The authors propose a locally adaptive choice of k that adapts simultaneously to the local Hölder exponent and intrinsic dimension, both potentially varying across the covariate domain.
- Extensive simulations back up the theory.
More from Research
- Voyager 1 Reaches One Light-Day From Earth, Signals Now Take 24 Hours — snowclipsed · 2026-11-18
- TCS Hiring Statement Urges Quality Over Quantity as AI Fuels Paper Arms Race — fortnow · 2026-11-18
- New paper: a sociocultural approach to linguistic creativity — abenitezburraco · 2026-11-18
- 27B Model Rewrites Its Own Weights Without Training or Gradients, Paper Claims — udmrzn · 2026-11-18
- From 'impossible' to a working Bluetooth beacon: open-source EDA chip design is real — IgorCarron · 2026-11-18
- AURA 2.0: a streaming video LLM that speaks first, calls tools, and grounds visually — Shuhan Li · 2026-11-18