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. 2402.15145
18
2

The Cost of Parallelizing Boosting

23 February 2024
Xin Lyu
Hongxun Wu
Junzhao Yang
ArXivPDFHTML
Abstract

We study the cost of parallelizing weak-to-strong boosting algorithms for learning, following the recent work of Karbasi and Larsen. Our main results are two-fold: - First, we prove a tight lower bound, showing that even "slight" parallelization of boosting requires an exponential blow-up in the complexity of training. Specifically, let γ\gammaγ be the weak learner's advantage over random guessing. The famous \textsc{AdaBoost} algorithm produces an accurate hypothesis by interacting with the weak learner for O~(1/γ2)\tilde{O}(1 / \gamma^2)O~(1/γ2) rounds where each round runs in polynomial time. Karbasi and Larsen showed that "significant" parallelization must incur exponential blow-up: Any boosting algorithm either interacts with the weak learner for Ω(1/γ)\Omega(1 / \gamma)Ω(1/γ) rounds or incurs an exp⁡(d/γ)\exp(d / \gamma)exp(d/γ) blow-up in the complexity of training, where ddd is the VC dimension of the hypothesis class. We close the gap by showing that any boosting algorithm either has Ω(1/γ2)\Omega(1 / \gamma^2)Ω(1/γ2) rounds of interaction or incurs a smaller exponential blow-up of exp⁡(d)\exp(d)exp(d). -Complementing our lower bound, we show that there exists a boosting algorithm using O~(1/(tγ2))\tilde{O}(1/(t \gamma^2))O~(1/(tγ2)) rounds, and only suffer a blow-up of exp⁡(d⋅t2)\exp(d \cdot t^2)exp(d⋅t2). Plugging in t=ω(1)t = \omega(1)t=ω(1), this shows that the smaller blow-up in our lower bound is tight. More interestingly, this provides the first trade-off between the parallelism and the total work required for boosting.

View on arXiv
Comments on this paper