More Like this

1. On Abruptly-Changing and Slowly-Varying Multiarmed Bandit Problems

   Wei, Lai; Srivastava, Vaibhav (June 2018, 2018 Annual American Control Conference)

   We study the non-stationary stochastic multi- armed bandit (MAB) problem and propose two generic algorithms, namely, Limited Memory Deterministic Sequencing of Exploration and Exploitation (LM-DSEE) and Sliding-Window Upper Confidence Bound# (SW-UCB#). We rigorously analyze these algorithms in abruptly-changing and slowly-varying environments and characterize their performance. We show that the expected cumulative regret for these algorithms in either of the environments is upper bounded by sublinear functions of time, i.e., the time average of the regret asymptotically converges to zero. We complement our analysis with numerical illustrations. 

   more » « less
2. Non-stationary Reinforcement Learning under General Function Approximation

   Feng, Songtao and; Yin, Ming; Huang, Ruiquan; Wang, Yu-Xiang; Yang, Jing; Liang, Yingbin (July 2023, Proceedings of Machine Learning Research)

   General function approximation is a powerful tool to handle large state and action spaces in a broad range of reinforcement learning (RL) scenarios. However, theoretical understanding of non-stationary MDPs with general function approximation is still limited. In this paper, we make the first such an attempt. We first propose a new complexity metric called dynamic Bellman Eluder (DBE) dimension for non-stationary MDPs, which subsumes majority of existing tractable RL problems in static MDPs as well as non-stationary MDPs. Based on the proposed complexity metric, we propose a novel confidence-set based model-free algorithm called SW-OPEA, which features a sliding window mechanism and a new confidence set design for non-stationary MDPs. We then establish an upper bound on the dynamic regret for the proposed algorithm, and show that SW-OPEA is provably efficient as long as the variation budget is not significantly large. We further demonstrate via examples of non-stationary linear and tabular MDPs that our algorithm performs better in small variation budget scenario than the existing UCB-type algorithms. To the best of our knowledge, this is the first dynamic regret analysis in non-stationary MDPs with general function approximation. 

   more » « less
3. Sliding-Window QPS (SW-QPS): A Perfect Parallel Iterative Switching Algorithm for Input-Queued Switches

   https://doi.org/10.1145/3453953.3453969  

   Meng, Jingfan; Gong, Long; Xu, Jun (March 2021, ACM SIGMETRICS Performance Evaluation Review)

   null (Ed.)

   In this work, we first propose a parallel batch switching algorithm called Small-Batch Queue-Proportional Sampling (SB-QPS). Compared to other batch switching algorithms, SB-QPS significantly reduces the batch size without sacrificing the throughput performance and hence has much lower delay when traffic load is light to moderate. It also achieves the lowest possible time complexity of O(1) per matching computation per port, via parallelization. We then propose another algorithm called Sliding-Window QPS (SW-QPS). SW-QPS retains and enhances all benefits of SB-QPS, and reduces the batching delay to zero via a novel switching framework called sliding-window switching. In addition, SW-QPS computes matchings of much higher qualities, as measured by the resulting throughput and delay performances, than QPS-1, the state-of-the-art regular switching algorithm that builds upon the same underlying bipartite matching algorithm. 

   more » « less
4. Multitask Bandit Learning Through Heterogeneous Feedback Aggregation

   Wang, Z.; Chaudhuri, K. (January 2021, Proceedings of the 24th International Conference on Artificial Intelligence and Statistics (AISTATS) 2021)

   In many real-world applications, multiple agents seek to learn how to perform highly related yet slightly different tasks in an online bandit learning protocol. We formulate this problem as the ϵ-multi-player multi-armed bandit problem, in which a set of players concurrently interact with a set of arms, and for each arm, the reward distributions for all players are similar but not necessarily identical. We develop an upper confidence bound-based algorithm, RobustAgg(ϵ), that adaptively aggregates rewards collected by different players. In the setting where an upper bound on the pairwise dissimilarities of reward distributions between players is known, we achieve instance-dependent regret guarantees that depend on the amenability of information sharing across players. We complement these upper bounds with nearly matching lower bounds. In the setting where pairwise dissimilarities are unknown, we provide a lower bound, as well as an algorithm that trades off minimax regret guarantees for adaptivity to unknown similarity structure. 

   more » « less
5. Finite-Time Frequentist Regret Bounds of Multi-Agent Thompson Sampling on Sparse Hypergraphs

   Jin, Tianyuan; Hsu, Hao-Lun; Chang, William; Xu, Pan (March 2024, The Thirty-Eighth AAAI Conference on Artificial Intelligence (AAAI))

   We study the multi-agent multi-armed bandit (MAMAB) problem, where agents are factored into overlapping groups. Each group represents a hyperedge, forming a hypergraph over the agents. At each round of interaction, the learner pulls a joint arm (composed of individual arms for each agent) and receives a reward according to the hypergraph structure. Specifically, we assume there is a local reward for each hyperedge, and the reward of the joint arm is the sum of these local rewards. Previous work introduced the multi-agent Thompson sampling (MATS) algorithm and derived a Bayesian regret bound. However, it remains an open problem how to derive a frequentist regret bound for Thompson sampling in this multi-agent setting. To address these issues, we propose an efficient variant of MATS, the epsilon-exploring Multi-Agent Thompson Sampling (eps-MATS) algorithm, which performs MATS exploration with probability epsilon while adopts a greedy policy otherwise. We prove that eps-MATS achieves a worst-case frequentist regret bound that is sublinear in both the time horizon and the local arm size. We also derive a lower bound for this setting, which implies our frequentist regret upper bound is optimal up to constant and logarithm terms, when the hypergraph is sufficiently sparse. Thorough experiments on standard MAMAB problems demonstrate the superior performance and the improved computational efficiency of eps-MATS compared with existing algorithms in the same setting. 

   more » « less