The exploration problem is one of the main challenges in deep reinforcement learning (RL). Recent promising works tried to handle the problem with population-based methods, which collect samples with diverse behaviors derived from a population of different exploratory policies. Adaptive policy selection has been adopted for behavior control. However, the behavior selection space is largely limited by the predefined policy population, which further limits behavior diversity. In this paper, we propose a general framework called Learnable Behavioral Control (LBC) to address the limitation, which a) enables a significantly enlarged behavior selection space via formulating a hybrid behavior mapping from all policies; b) constructs a unified learnable process for behavior selection. We introduce LBC into distributed off-policy actor-critic methods and achieve behavior control via optimizing the selection of the behavior mappings with bandit-based meta-controllers. Our agents have achieved 10077.52% mean human normalized score and surpassed 24 human world records within 1B training frames in the Arcade Learning Environment, which demonstrates our significant state-of-the-art (SOTA) performance without degrading the sample efficiency.

A method using deep reinforcement learning (DRL) to non-iteratively generate an optimal mesh for an arbitrary blade passage is developed. Despite automation in mesh generation using either an empirical approach or an optimization algorithm, repeated tuning of meshing parameters is still required for a new geometry. The method developed herein employs a DRL-based multi-condition optimization technique to define optimal meshing parameters as a function of the blade geometry, attaining automation, minimization of human intervention, and computational efficiency. The meshing parameters are optimized by training an elliptic mesh generator which generates a structured mesh for a blade passage with an arbitrary blade geometry. During each episode of the DRL process, the mesh generator is trained to produce an optimal mesh for a randomly selected blade passage by updating the meshing parameters until the mesh quality, as measured by the ratio of determinants of the Jacobian matrices and the skewness, reaches the highest level. Once the training is completed, the mesh generator create an optimal mesh for a new arbitrary blade passage in a single try without an repetitive process for the parameter tuning for mesh generation from the scratch. The effectiveness and robustness of the proposed method are demonstrated through the generation of meshes for various blade passages.

Standard approaches to sequential decision-making exploit an agent's ability to continually interact with its environment and improve its control policy. However, due to safety, ethical, and practicality constraints, this type of trial-and-error experimentation is often infeasible in many real-world domains such as healthcare and robotics. Instead, control policies in these domains are typically trained offline from previously logged data or in a growing-batch manner. In this setting a fixed policy is deployed to the environment and used to gather an entire batch of new data before being aggregated with past batches and used to update the policy. This improvement cycle can then be repeated multiple times. While a limited number of such cycles is feasible in real-world domains, the quality and diversity of the resulting data are much lower than in the standard continually-interacting approach. However, data collection in these domains is often performed in conjunction with human experts, who are able to label or annotate the collected data. In this paper, we first explore the trade-offs present in this growing-batch setting, and then investigate how information provided by a teacher (i.e., demonstrations, expert actions, and gradient information) can be leveraged at training time to mitigate the sample complexity and coverage requirements for actor-critic methods. We validate our contributions on tasks from the DeepMind Control Suite.

There is a lack of standard benchmarks for Multi-Agent Reinforcement Learning (MARL) algorithms. The Starcraft Multi-Agent Challenge (SMAC) has been widely used in MARL research, but is built on top of a heavy, closed-source computer game, StarCraft II. Thus, SMAC is computationally expensive and requires knowledge and the use of proprietary tools specific to the game for any meaningful alteration or contribution to the environment. We introduce SMAClite -- a challenge based on SMAC that is both decoupled from Starcraft II and open-source, along with a framework which makes it possible to create new content for SMAClite without any special knowledge. We conduct experiments to show that SMAClite is equivalent to SMAC, by training MARL algorithms on SMAClite and reproducing SMAC results. We then show that SMAClite outperforms SMAC in both runtime speed and memory.

Reinforcement learning on high-dimensional and complex problems relies on abstraction for improved efficiency and generalization. In this paper, we study abstraction in the continuous-control setting, and extend the definition of MDP homomorphisms to the setting of continuous state and action spaces. We derive a policy gradient theorem on the abstract MDP for both stochastic and deterministic policies. Our policy gradient results allow for leveraging approximate symmetries of the environment for policy optimization. Based on these theorems, we propose a family of actor-critic algorithms that are able to learn the policy and the MDP homomorphism map simultaneously, using the lax bisimulation metric. Finally, we introduce a series of environments with continuous symmetries to further demonstrate the ability of our algorithm for action abstraction in the presence of such symmetries. We demonstrate the effectiveness of our method on our environments, as well as on challenging visual control tasks from the DeepMind Control Suite. Our method's ability to utilize MDP homomorphisms for representation learning leads to improved performance, and the visualizations of the latent space clearly demonstrate the structure of the learned abstraction.

Reinforcement learning (RL) problems over general state and action spaces are notoriously challenging. In contrast to the tableau setting, one can not enumerate all the states and then iteratively update the policies for each state. This prevents the application of many well-studied RL methods especially those with provable convergence guarantees. In this paper, we first present a substantial generalization of the recently developed policy mirror descent method to deal with general state and action spaces. We introduce new approaches to incorporate function approximation into this method, so that we do not need to use explicit policy parameterization at all. Moreover, we present a novel policy dual averaging method for which possibly simpler function approximation techniques can be applied. We establish linear convergence rate to global optimality or sublinear convergence to stationarity for these methods applied to solve different classes of RL problems under exact policy evaluation. We then define proper notions of the approximation errors for policy evaluation and investigate their impact on the convergence of these methods applied to general-state RL problems with either finite-action or continuous-action spaces. To the best of our knowledge, the development of these algorithmic frameworks as well as their convergence analysis appear to be new in the literature.

This study explores the vaccine prioritization strategy to reduce the overall burden of the pandemic when the supply is limited. Existing methods conduct macro-level or simplified micro-level vaccine distribution by assuming the homogeneous behavior within subgroup populations and lacking mobility dynamics integration. Directly applying these models for micro-level vaccine allocation leads to sub-optimal solutions due to the lack of behavioral-related details. To address the issue, we first incorporate the mobility heterogeneity in disease dynamics modeling and mimic the disease evolution process using a Trans-vaccine-SEIR model. Then we develop a novel deep reinforcement learning to seek the optimal vaccine allocation strategy for the high-degree spatial-temporal disease evolution system. The graph neural network is used to effectively capture the structural properties of the mobility contact network and extract the dynamic disease features. In our evaluation, the proposed framework reduces 7% - 10% of infections and deaths than the baseline strategies. Extensive evaluation shows that the proposed framework is robust to seek the optimal vaccine allocation with diverse mobility patterns in the micro-level disease evolution system. In particular, we find the optimal vaccine allocation strategy in the transit usage restriction scenario is significantly more effective than restricting cross-zone mobility for the top 10% age-based and income-based zones. These results provide valuable insights for areas with limited vaccines and low logistic efficacy.

This paper presents a novel algorithm for the continuous control of dynamical systems that combines Trajectory Optimization (TO) and Reinforcement Learning (RL) in a single framework. The motivations behind this algorithm are the two main limitations of TO and RL when applied to continuous nonlinear systems to minimize a non-convex cost function. Specifically, TO can get stuck in poor local minima when the search is not initialized close to a "good" minimum. On the other hand, when dealing with continuous state and control spaces, the RL training process may be excessively long and strongly dependent on the exploration strategy. Thus, our algorithm learns a "good" control policy via TO-guided RL policy search that, when used as initial guess provider for TO, makes the trajectory optimization process less prone to converge to poor local optima. Our method is validated on several reaching problems featuring non-convex obstacle avoidance with different dynamical systems, including a car model with 6D state, and a 3-joint planar manipulator. Our results show the great capabilities of CACTO in escaping local minima, while being more computationally efficient than the Deep Deterministic Policy Gradient (DDPG) and Proximal Policy Optimization (PPO) RL algorithms.

Policy optimization methods with function approximation are widely used in multi-agent reinforcement learning. However, it remains elusive how to design such algorithms with statistical guarantees. Leveraging a multi-agent performance difference lemma that characterizes the landscape of multi-agent policy optimization, we find that the localized action value function serves as an ideal descent direction for each local policy. Motivated by the observation, we present a multi-agent PPO algorithm in which the local policy of each agent is updated similarly to vanilla PPO. We prove that with standard regularity conditions on the Markov game and problem-dependent quantities, our algorithm converges to the globally optimal policy at a sublinear rate. We extend our algorithm to the off-policy setting and introduce pessimism to policy evaluation, which aligns with experiments. To our knowledge, this is the first provably convergent multi-agent PPO algorithm in cooperative Markov games.

Model-based reinforcement learning (MBRL) techniques have recently yielded promising results for real-world autonomous racing using high-dimensional observations. MBRL agents, such as Dreamer, solve long-horizon tasks by building a world model and planning actions by latent imagination. This approach involves explicitly learning a model of the system dynamics and using it to learn the optimal policy for continuous control over multiple timesteps. As a result, MBRL agents may converge to sub-optimal policies if the world model is inaccurate. To improve state estimation for autonomous racing, this paper proposes a self-supervised sensor fusion technique that combines egocentric LiDAR and RGB camera observations collected from the F1TENTH Gym. The zero-shot performance of MBRL agents is empirically evaluated on unseen tracks and against a dynamic obstacle. This paper illustrates that multimodal perception improves robustness of the world model without requiring additional training data. The resulting multimodal Dreamer agent safely avoided collisions and won the most races compared to other tested baselines in zero-shot head-to-head autonomous racing.