A Two-Timescale Framework for Bilevel Optimization: Complexity Analysis and Application to Actor-Critic
Preprint
Abstract
This paper analyzes a two-timescale stochastic algorithm framework for bilevel optimization. Bilevel optimization is a class of problems which exhibit a two-level structure, and its goal is to minimize an outer objective function with variables which are constrained to be the optimal solution to an (inner) optimization problem. We consider the case when the inner problem is unconstrained and strongly convex, while the outer problem is constrained and has a smooth objective function. We propose a two-timescale stochastic approximation (TTSA) algorithm for tackling such a bilevel problem. In the algorithm, a stochastic gradient update with a larger step size is used for the inner problem, while a projected stochastic gradient update with a smaller step size is used for the outer problem. We analyze the convergence rates for the TTSA algorithm under various settings: when the outer problem is strongly convex (resp.~weakly convex), the TTSA algorithm finds an $\mathcal{O}(K^{-2/3})$-optimal (resp.~$\mathcal{O}(K^{-2/5})$-stationary) solution, where $K$ is the total iteration number. As an application, we show that a two-timescale natural actor-critic proximal policy optimization algorithm can be viewed as a special case of our TTSA framework. Importantly, the natural actor-critic algorithm is shown to converge at a rate of $\mathcal{O}(K^{-1/4})$ in terms of the gap in expected discounted reward compared to a global optimal policy.
Keywords
All Related Versions
This publication has 16 references indexed in Scilit:
- A Two-Timescale Stochastic Algorithm Framework for Bilevel Optimization: Complexity Analysis and Application to Actor-CriticSIAM Journal on Optimization, 2023
- A First Order Method for Solving Convex Bilevel Optimization ProblemsSIAM Journal on Optimization, 2017
- Stochastic compositional gradient descent: algorithms for minimizing compositions of expected-value functionsMathematical Programming, 2016
- Natural Actor-CriticNeurocomputing, 2008
- Learning near-optimal policies with Bellman-residual minimization based fitted policy iteration and a single sample pathMachine Learning, 2007
- Convergence rate and averaging of nonlinear two-time-scale stochastic approximation algorithmsThe Annals of Applied Probability, 2006
- On bilevel programming, Part I: General nonlinear casesMathematical Programming, 1995
- Descent approaches for quadratic bilevel programmingJournal of Optimization Theory and Applications, 1994
- A penalty function approach for solving bi-level linear programsJournal of Global Optimization, 1993
- Double penalty method for bilevel optimization problemsAnnals of Operations Research, 1992