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. 2207.13082
17
6

Offline Reinforcement Learning at Multiple Frequencies

26 July 2022
Kaylee Burns
Tianhe Yu
Chelsea Finn
Karol Hausman
    OffRL
ArXivPDFHTML
Abstract

Leveraging many sources of offline robot data requires grappling with the heterogeneity of such data. In this paper, we focus on one particular aspect of heterogeneity: learning from offline data collected at different control frequencies. Across labs, the discretization of controllers, sampling rates of sensors, and demands of a task of interest may differ, giving rise to a mixture of frequencies in an aggregated dataset. We study how well offline reinforcement learning (RL) algorithms can accommodate data with a mixture of frequencies during training. We observe that the QQQ-value propagates at different rates for different discretizations, leading to a number of learning challenges for off-the-shelf offline RL. We present a simple yet effective solution that enforces consistency in the rate of QQQ-value updates to stabilize learning. By scaling the value of NNN in NNN-step returns with the discretization size, we effectively balance QQQ-value propagation, leading to more stable convergence. On three simulated robotic control problems, we empirically find that this simple approach outperforms na\"ive mixing by 50% on average.

View on arXiv
Comments on this paper