ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 1712.02369
40
13
v1v2 (latest)

Achieving the time of 111-NN, but the accuracy of kkk-NN

6 December 2017
Lirong Xue
Samory Kpotufe
ArXiv (abs)PDFHTML
Abstract

We propose a simple approach which, given distributed computing resources, can nearly achieve the accuracy of kkk-NN prediction, while matching (or improving) the faster prediction time of 111-NN. The approach consists of aggregating denoised 111-NN predictors over a small number of distributed subsamples. We show, both theoretically and experimentally, that small subsample sizes suffice to attain similar performance as kkk-NN, without sacrificing the computational efficiency of 111-NN.

View on arXiv
Comments on this paper