More Like this

1. Posterior sampling with delayed feedback for reinforcement learning with linear function approximation

   Kuang, Nikki Lijing; Yin, Ming; Wang, Mengdi; Wang, Yu-Xiang; Ma, Yi-An (December 2023, Proceedings of Machine Learning Research)

   Recent studies in reinforcement learning (RL) have made significant progress by leveraging function approximation to alleviate the sample complexity hurdle for better performance. Despite the success, existing provably efficient algorithms typically rely on the accessibility of immediate feedback upon taking actions. The failure to account for the impact of delay in observations can significantly degrade the performance of real-world systems due to the regret blow-up. In this work, we tackle the challenge of delayed feedback in RL with linear function approximation by employing posterior sampling, which has been shown to empirically outperform the popular UCB algorithms in a wide range of regimes. We first introduce Delayed-PSVI, an optimistic value-based algorithm that effectively explores the value function space via noise perturbation with posterior sampling. We provide the first analysis for posterior sampling algorithms with delayed feedback in RL and show our algorithm achieves $$\widetilde{O}(\sqrt{d^3H^3 T} + d^2H^2 E[\tau])$$ worst-case regret in the presence of unknown stochastic delays. Here $$E[\tau]$$ is the expected delay. To further improve its computational efficiency and to expand its applicability in high-dimensional RL problems, we incorporate a gradient-based approximate sampling scheme via Langevin dynamics for Delayed-LPSVI, which maintains the same order-optimal regret guarantee with $$\widetilde{O}(dHK)$$ computational cost. Empirical evaluations are performed to demonstrate the statistical and computational efficacy of our algorithms. 

   more » « less
2. Link rate selection using constrained Thompson sampling

   Gupta, H; Eryilmaz, A; Srikant, R (April 2019, IEEE INFOCOM)

   We consider the optimal link rate selection problem in time-varying wireless channels with unknown channel statistics. The aim of optimal link rate selection is to transmit at the optimal rate at each time slot in order to maximize the expected throughput of the wireless channel/link or equivalently minimize the expected regret. Lack of information about channel state or channel statistics necessitates the use of online/sequential learning algorithms to determine the optimal rate. We present an algorithm called CoTS - Constrained Thompson sampling algorithm which improves upon the current state-of-the-art, is fast and is also general in the sense that it can handle several different constraints in the problem with the same algorithm. We also prove an asymptotic lower bound on the expected regret and a high probability large-horizon upper bound on the regret, which show that the regret grows logarithmically with time in an order sense. We also provide numerical results which establish that CoTS significantly outperforms the current state-of-the-art algorithms. 

   more » « less
3. No-Regret Linear Bandits beyond Realizability

   Liu, Chong and; Yin, Ming; Wang, Yu-Xiang (July 2023, Proceedings of Machine Learning Research)

   Evans, Robin J.; Shpitser, Ilya (Ed.)

   We study linear bandits when the underlying reward function is not linear. Existing work relies on a uniform misspecification parameter $$\epsilon$$ that measures the sup-norm error of the best linear approximation. This results in an unavoidable linear regret whenever $$\epsilon > 0$$. We describe a more natural model of misspecification which only requires the approximation error at each input $$x$$ to be proportional to the suboptimality gap at $$x$$. It captures the intuition that, for optimization problems, near-optimal regions should matter more and we can tolerate larger approximation errors in suboptimal regions. Quite surprisingly, we show that the classical LinUCB algorithm — designed for the realizable case — is automatically robust against such gap-adjusted misspecification. It achieves a near-optimal $$\sqrt{T}$$ regret for problems that the best-known regret is almost linear in time horizon $$T$$. Technically, our proof relies on a novel self-bounding argument that bounds the part of the regret due to misspecification by the regret itself. 

   more » « less
4. Optimal Batched Linear Bandits

   Ren, Xuanfei; Jin, Tianyuan; Xu, Pan (May 2024, Proceedings of The 41st International Conference on Machine Learning (ICML))

   We introduce the E$^4$ algorithm for the batched linear bandit problem, incorporating an Explore-Estimate-Eliminate-Exploit framework. With a proper choice of exploration rate, we prove E$^4$ achieves the finite-time minimax optimal regret with only $$O(\log\log T)$$ batches, and the asymptotically optimal regret with only $$3$$ batches as $$T\rightarrow\infty$$, where $$T$$ is the time horizon. We further prove a lower bound on the batch complexity of linear contextual bandits showing that any asymptotically optimal algorithm must require at least $$3$$ batches in expectation as $$T\rightarrow\infty$$, which indicates E$^4$ achieves the asymptotic optimality in regret and batch complexity simultaneously. To the best of our knowledge, E$^4$ is the first algorithm for linear bandits that simultaneously achieves the minimax and asymptotic optimality in regret with the corresponding optimal batch complexities. In addition, we show that with another choice of exploration rate E$^4$ achieves an instance-dependent regret bound requiring at most $$O(\log T)$$ batches, and maintains the minimax optimality and asymptotic optimality. We conduct thorough experiments to evaluate our algorithm on randomly generated instances and the challenging \textit{End of Optimism} instances \citep{lattimore2017end} which were shown to be hard to learn for optimism based algorithms. Empirical results show that E$^4$ consistently outperforms baseline algorithms with respect to regret minimization, batch complexity, and computational efficiency. 

   more » « less
5. Optimal Batched Linear Bandits

   Ren, Xuanfei; Jin, Tianyuan; Xu, Pan (May 2024, Proceedings of the 41st International Conference on Machine Learning)

   We introduce the E$^4$ algorithm for the batched linear bandit problem, incorporating an Explore-Estimate-Eliminate-Exploit framework. With a proper choice of exploration rate, we prove E$^4$ achieves the finite-time minimax optimal regret with only $$O(\log\log T)$$ batches, and the asymptotically optimal regret with only $$3$$ batches as $$T\rightarrow\infty$$, where $$T$$ is the time horizon. We further prove a lower bound on the batch complexity of linear contextual bandits showing that any asymptotically optimal algorithm must require at least $$3$$ batches in expectation as $$T\rightarrow\infty$$, which indicates E$^4$ achieves the asymptotic optimality in regret and batch complexity simultaneously. To the best of our knowledge, E$^4$ is the first algorithm for linear bandits that simultaneously achieves the minimax and asymptotic optimality in regret with the corresponding optimal batch complexities. In addition, we show that with another choice of exploration rate E$^4$ achieves an instance-dependent regret bound requiring at most $$O(\log T)$$ batches, and maintains the minimax optimality and asymptotic optimality. We conduct thorough experiments to evaluate our algorithm on randomly generated instances and the challenging \textit{End of Optimism} instances \citep{lattimore2017end} which were shown to be hard to learn for optimism based algorithms. Empirical results show that E$^4$ consistently outperforms baseline algorithms with respect to regret minimization, batch complexity, and computational efficiency. 

   more » « less