English

Learning Smooth Distance Functions via Queries

Machine Learning 2024-12-03 v1 Artificial Intelligence Information Retrieval Machine Learning

Abstract

In this work, we investigate the problem of learning distance functions within the query-based learning framework, where a learner is able to pose triplet queries of the form: ``Is xix_i closer to xjx_j or xkx_k?'' We establish formal guarantees on the query complexity required to learn smooth, but otherwise general, distance functions under two notions of approximation: ω\omega-additive approximation and (1+ω)(1 + \omega)-multiplicative approximation. For the additive approximation, we propose a global method whose query complexity is quadratic in the size of a finite cover of the sample space. For the (stronger) multiplicative approximation, we introduce a method that combines global and local approaches, utilizing multiple Mahalanobis distance functions to capture local geometry. This method has a query complexity that scales quadratically with both the size of the cover and the ambient space dimension of the sample space.

Keywords

Cite

@article{arxiv.2412.01290,
  title  = {Learning Smooth Distance Functions via Queries},
  author = {Akash Kumar and Sanjoy Dasgupta},
  journal= {arXiv preprint arXiv:2412.01290},
  year   = {2024}
}

Comments

40 pages, 1 figure

R2 v1 2026-06-28T20:19:22.956Z