TwoTimescale Stochastic Approximation Convergence Rates with Applications to Reinforcement Learning
This topic contains 0 replies, has 1 voice, and was last updated by arXiv 2 months, 2 weeks ago.

TwoTimescale Stochastic Approximation Convergence Rates with Applications to Reinforcement Learning
Twotimescale Stochastic Approximation (SA) algorithms are widely used in Reinforcement Learning (RL). Their iterates have two parts that are updated with distinct stepsizes. In this work we provide a recipe for analyzing twotimescale SA. Using it, we develop the first convergence rate result for them. From this result we extract key insights on stepsize selection. As an application, we obtain convergence rates for twotimescale RL algorithms such as GTD(0), GTD2, and TDC.
TwoTimescale Stochastic Approximation Convergence Rates with Applications to Reinforcement Learning
by Gal Dalal, Balazs Szorenyi, Gugan Thoppe, Shie Mannor
https://arxiv.org/pdf/1703.05376v3.pdf
You must be logged in to reply to this topic.