A Tale of Two-Timescale Reinforcement Learning with the Tightest Finite-Time Bound
This provides a rigorous finite-time analysis for policy evaluation in RL, offering tighter bounds and broader applicability than prior work, which is incremental but important for algorithm design and theory.
The paper tackles the problem of analyzing convergence rates for two-timescale reinforcement learning algorithms like GTD methods, showing that with specific stepsize sequences, the iterates converge at tight rates of O~(n^{-α/2}) and O~(n^{-β/2}) with high probability, and proving these bounds are optimal via lower bounds.
Policy evaluation in reinforcement learning is often conducted using two-timescale stochastic approximation, which results in various gradient temporal difference methods such as GTD(0), GTD2, and TDC. Here, we provide convergence rate bounds for this suite of algorithms. Algorithms such as these have two iterates, $θ_n$ and $w_n,$ which are updated using two distinct stepsize sequences, $α_n$ and $β_n,$ respectively. Assuming $α_n = n^{-α}$ and $β_n = n^{-β}$ with $1 > α> β> 0,$ we show that, with high probability, the two iterates converge to their respective solutions $θ^*$ and $w^*$ at rates given by $\|θ_n - θ^*\| = \tilde{O}( n^{-α/2})$ and $\|w_n - w^*\| = \tilde{O}(n^{-β/2});$ here, $\tilde{O}$ hides logarithmic terms. Via comparable lower bounds, we show that these bounds are, in fact, tight. To the best of our knowledge, ours is the first finite-time analysis which achieves these rates. While it was known that the two timescale components decouple asymptotically, our results depict this phenomenon more explicitly by showing that it in fact happens from some finite time onwards. Lastly, compared to existing works, our result applies to a broader family of stepsizes, including non-square summable ones.