More Like this

1. Value-Decomposition Multi-Agent Actor-Critics

   Su, Jianyu; Adams, Stephen; Beling, Peter (February 2021, Proceedings of the AAAI Conference on Artificial Intelligence)

   The exploitation of extra state information has been an active research area in multi-agent reinforcement learning (MARL). QMIX represents the joint action-value using a non-negative function approximator and achieves the best performance on the StarCraft II micromanagement testbed, a common MARL benchmark. However, our experiments demonstrate that, in some cases, QMIX performs sub-optimally with the A2C framework, a training paradigm that promotes algorithm training efficiency. To obtain a reasonable trade-off between training efficiency and algorithm performance, we extend value-decomposition to actor-critic methods that are compatible with A2C and propose a novel actor-critic framework, value-decomposition actor-critic (VDAC). We evaluate VDAC on the StarCraft II micromanagement task and demonstrate that the proposed framework improves median performance over other actor-critic methods. Furthermore, we use a set of ablation experiments to identify the key factors that contribute to the performance of VDAC. 

   more » « less
2. Value-decomposition multi-agent actor-critics

   Su, Jianyu; Adams, Stephen; Beling, Peter (February 2021, Proceedings of the AAAI Conference on Artificial Intelligence)

   The exploitation of extra state information has been an active research area in multi-agent reinforcement learning (MARL). QMIX represents the joint action-value using a non-negative function approximator and achieves the best performance on the StarCraft II micromanagement testbed, a common MARL benchmark. However, our experiments demonstrate that, in some cases, QMIX performs sub-optimally with the A2C framework, a training paradigm that promotes algorithm training efficiency. To obtain a reasonable trade-off between training efficiency and algorithm performance, we extend value-decomposition to actor-critic methods that are compatible with A2C and propose a novel actor-critic framework, value-decomposition actor-critic (VDAC). We evaluate VDAC on the StarCraft II micromanagement task and demonstrate that the proposed framework improves median performance over other actor-critic methods. Furthermore, we use a set of ablation experiments to identify the key factors that contribute to the performance of VDAC. 

   more » « less
3. Unbiased Asymmetric Reinforcement Learning under Partial Observability

   Baisero, Andrea; Amato, Christopher (January 2022, Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems)

   In partially observable reinforcement learning, offline training gives access to latent information which is not available during online training and/or execution, such as the system state. Asymmetric actor-critic methods exploit such information by training a history-based policy via a state-based critic. However, many asymmetric methods lack theoretical foundation, and are only evaluated on limited domains. We examine the theory of asymmetric actor-critic methods which use state-based critics, and expose fundamental issues which undermine the validity of a common variant, and limit its ability to address partial observability. We propose an unbiased asymmetric actor-critic variant which is able to exploit state information while remaining theoretically sound, maintaining the validity of the policy gradient theorem, and introducing no bias and relatively low variance into the training process. An empirical evaluation performed on domains which exhibit significant partial observability confirms our analysis, demonstrating that unbiased asymmetric actor-critic converges to better policies and/or faster than symmetric and biased asymmetric baselines. 

   more » « less
4. Stochastic Approximation with Unbounded Markovian Noise: A General-Purpose Theorem

   Haque, Shaan Ul; Maguluri, Siva Theja (May 2025, AISTATS)

   Motivated by engineering applications such as resource allocation in networks and inventory systems, we consider average-reward Reinforcement Learning with unbounded state space and reward function. Recent work Murthy et al. (2024) studied this problem in the actor-critic framework and established finite sample bounds assuming access to a critic with certain error guarantees. We complement their work by studying Temporal Difference (TD) learning with linear function approximation and establishing finite-time bounds with the optimal sample complexity. These results are obtained using the following general-purpose theorem for non-linear Stochastic Approximation (SA). Suppose that one constructs a Lyapunov function for a non-linear SA with certain drift condition. Then, our theorem establishes finite-time bounds when this SA is driven by unbounded Markovian noise under suitable conditions. It serves as a black box tool to generalize sample guarantees on SA from i.i.d. or martingale difference case to potentially unbounded Markovian noise. The generality and the mild assumptions of the setup enables broad applicability of our theorem. We illustrate its power by studying two more systems: (i) We improve upon the finite-time bounds of Q-learning in Chen et al. (2024) by tightening the error bounds and also allowing for a larger class of behavior policies. (ii) We establish the first ever finite-time bounds for distributed stochastic optimization of high-dimensional smooth strongly convex function using cyclic block coordinate descent. 

   more » « less
5. Safe reinforcement learning for dynamical games

   https://doi.org/10.1002/rnc.4962  

   Yang, Yongliang; Vamvoudakis, Kyriakos G.; Modares, Hamidreza (March 2020, International Journal of Robust and Nonlinear Control)

   Summary This article presents a novel actor‐critic‐barrier structure for the multiplayer safety‐critical systems. Non‐zero‐sum (NZS) games with full‐state constraints are first transformed into unconstrained NZS games using a barrier function. The barrier function is capable of dealing with both symmetric and asymmetric constraints on the state. It is shown that the Nash equilibrium of the unconstrained NZS guarantees to stabilize the original multiplayer system. The barrier function is combined with an actor‐critic structure to learn the Nash equilibrium solution in an online fashion. It is shown that integrating the barrier function with the actor‐critic structure guarantees that the constraints will not be violated during learning. Boundedness and stability of the closed‐loop signals are analyzed. The efficacy of the presented approach is finally demonstrated by using a simulation example. 

   more » « less