The financial market is a particularly challenging playground for deep reinforcement learning due to its unique feature of dynamic datasets. Building high-quality market environments for training financial reinforcement learning (FinRL) agents is difficult due to major factors such as the low signal-to-noise ratio of financial data, survivorship bias of historical data, and model overfitting. In this paper, we present FinRL-Meta, a data-centric and openly accessible library that processes dynamic datasets from real-world markets into gym-style market environments and has been actively maintained by the AI4Finance community. First, following a DataOps paradigm, we provide hundreds of market environments through an automatic data curation pipeline. Second, we provide homegrown examples and reproduce popular research papers as stepping stones for users to design new trading strategies. We also deploy the library on cloud platforms so that users can visualize their own results and assess the relative performance via community-wise competitions. Third, we provide dozens of Jupyter/Python demos organized into a curriculum and a documentation website to serve the rapidly growing community. The open-source codes for the data curation pipeline are available at https://github.com/AI4Finance-Foundation/FinRL-Meta

Treatment planning for chronic diseases is a critical task in medical artificial intelligence, particularly in traditional Chinese medicine (TCM). However, generating optimized sequential treatment strategies for patients with chronic diseases in different clinical encounters remains a challenging issue that requires further exploration. In this study, we proposed a TCM herbal prescription planning framework based on deep reinforcement learning for chronic disease treatment (PrescDRL). PrescDRL is a sequential herbal prescription optimization model that focuses on long-term effectiveness rather than achieving maximum reward at every step, thereby ensuring better patient outcomes. We constructed a high-quality benchmark dataset for sequential diagnosis and treatment of diabetes and evaluated PrescDRL against this benchmark. Our results showed that PrescDRL achieved a higher curative effect, with the single-step reward improving by 117% and 153% compared to doctors. Furthermore, PrescDRL outperformed the benchmark in prescription prediction, with precision improving by 40.5% and recall improving by 63%. Overall, our study demonstrates the potential of using artificial intelligence to improve clinical intelligent diagnosis and treatment in TCM.

The process of revising (or constructing) a policy immediately prior to execution -- known as decision-time planning -- is key to achieving superhuman performance in perfect-information settings like chess and Go. A recent line of work has extended decision-time planning to more general imperfect-information settings, leading to superhuman performance in poker. However, these methods requires considering subgames whose sizes grow quickly in the amount of non-public information, making them unhelpful when the amount of non-public information is large. Motivated by this issue, we introduce an alternative framework for decision-time planning that is not based on subgames but rather on the notion of update equivalence. In this framework, decision-time planning algorithms simulate updates of synchronous learning algorithms. This framework enables us to introduce a new family of principled decision-time planning algorithms that do not rely on public information, opening the door to sound and effective decision-time planning in settings with large amounts of non-public information. In experiments, members of this family produce comparable or superior results compared to state-of-the-art approaches in Hanabi and improve performance in 3x3 Abrupt Dark Hex and Phantom Tic-Tac-Toe.

Internet of Things (IoT) networks have become ubiquitous as autonomous computing, communication and collaboration among devices become popular for accomplishing various tasks. The use of relays in IoT networks further makes it convenient to deploy IoT networks as relays provide a host of benefits, like increasing the communication range and minimizing power consumption. Existing literature on traditional AoI schedulers for such two-hop relayed IoT networks are limited because they are designed assuming constant/non-changing channel conditions and known (usually, generate-at-will) packet generation patterns. Deep reinforcement learning (DRL) algorithms have been investigated for AoI scheduling in two-hop IoT networks with relays, however, they are only applicable for small-scale IoT networks due to exponential rise in action space as the networks become large. These limitations discourage the practical utilization of AoI schedulers for IoT network deployments. This paper presents a practical AoI scheduler for two-hop IoT networks with relays that addresses the above limitations. The proposed scheduler utilizes a novel voting mechanism based proximal policy optimization (v-PPO) algorithm that maintains a linear action space, enabling it be scale well with larger IoT networks. The proposed v-PPO based AoI scheduler adapts well to changing network conditions and accounts for unknown traffic generation patterns, making it practical for real-world IoT deployments. Simulation results show that the proposed v-PPO based AoI scheduler outperforms both ML and traditional (non-ML) AoI schedulers, such as, Deep Q Network (DQN)-based AoI Scheduler, Maximal Age First-Maximal Age Difference (MAF-MAD), MAF (Maximal Age First) , and round-robin in all considered practical scenarios.

Generalizability and stability are two key objectives for operating reinforcement learning (RL) agents in the real world. Designing RL algorithms that optimize these objectives can be a costly and painstaking process. This paper presents MetaPG, an evolutionary method for automated design of actor-critic loss functions. MetaPG explicitly optimizes for generalizability and performance, and implicitly optimizes the stability of both metrics. We initialize our loss function population with Soft Actor-Critic (SAC) and perform multi-objective optimization using fitness metrics encoding single-task performance, zero-shot generalizability to unseen environment configurations, and stability across independent runs with different random seeds. On a set of continuous control tasks from the Real-World RL Benchmark Suite, we find that our method, using a single environment during evolution, evolves algorithms that improve upon SAC's performance and generalizability by 4% and 20%, respectively, and reduce instability up to 67%. Then, we scale up to more complex environments from the Brax physics simulator and replicate generalizability tests encountered in practical settings, such as different friction coefficients. MetaPG evolves algorithms that can obtain 10% better generalizability without loss of performance within the same meta-training environment and obtain similar results to SAC when doing cross-domain evaluations in other Brax environments. The evolution results are interpretable; by analyzing the structure of the best algorithms we identify elements that help optimizing certain objectives, such as regularization terms for the critic loss.

Variational quantum algorithms are the leading candidate for advantage on near-term quantum hardware. When training a parametrized quantum circuit in this setting to solve a specific problem, the choice of ansatz is one of the most important factors that determines the trainability and performance of the algorithm. In quantum machine learning (QML), however, the literature on ansatzes that are motivated by the training data structure is scarce. In this work, we introduce an ansatz for learning tasks on weighted graphs that respects an important graph symmetry, namely equivariance under node permutations. We evaluate the performance of this ansatz on a complex learning task, namely neural combinatorial optimization, where a machine learning model is used to learn a heuristic for a combinatorial optimization problem. We analytically and numerically study the performance of our model, and our results strengthen the notion that symmetry-preserving ansatzes are a key to success in QML.

This paper investigates policy resilience to training-environment poisoning attacks on reinforcement learning (RL) policies, with the goal of recovering the deployment performance of a poisoned RL policy. Due to the fact that the policy resilience is an add-on concern to RL algorithms, it should be resource-efficient, time-conserving, and widely applicable without compromising the performance of RL algorithms. This paper proposes such a policy-resilience mechanism based on an idea of knowledge sharing. We summarize the policy resilience as three stages: preparation, diagnosis, recovery. Specifically, we design the mechanism as a federated architecture coupled with a meta-learning manner, pursuing an efficient extraction and sharing of the environment knowledge. With the shared knowledge, a poisoned agent can quickly identify the deployment condition and accordingly recover its policy performance. We empirically evaluate the resilience mechanism for both model-based and model-free RL algorithms, showing its effectiveness and efficiency in restoring the deployment performance of a poisoned policy.

We consider the development of adaptive, instance-dependent algorithms for interactive decision making (bandits, reinforcement learning, and beyond) that, rather than only performing well in the worst case, adapt to favorable properties of real-world instances for improved performance. We aim for instance-optimality, a strong notion of adaptivity which asserts that, on any particular problem instance, the algorithm under consideration outperforms all consistent algorithms. Instance-optimality enjoys a rich asymptotic theory originating from the work of \citet{lai1985asymptotically,graves1997asymptotically}, but non-asymptotic guarantees have remained elusive outside of certain special cases. Even for problems as simple as tabular reinforcement learning, existing algorithms do not attain instance-optimal performance until the number of rounds of interaction is doubly exponential in the number of states. In this paper, we take the first step toward developing a non-asymptotic theory of instance-optimal decision making with general function approximation. We introduce a new complexity measure, the Allocation-Estimation Coefficient (AEC), and provide a new algorithm, $\mathsf{AE}^2$, which attains non-asymptotic instance-optimal performance at a rate controlled by the AEC. Our results recover the best known guarantees for well-studied problems such as finite-armed and linear bandits and, when specialized to tabular reinforcement learning, attain the first instance-optimal regret bounds with polynomial dependence on all problem parameters, improving over prior work exponentially. We complement these results with lower bounds that show that i) existing notions of statistical complexity are insufficient to derive non-asymptotic guarantees, and ii) under certain technical conditions, boundedness of the AEC is necessary to learn an instance-optimal allocation of decisions in finite time.

We propose a model-free reinforcement learning solution, namely the ASAP-Phi framework, to encourage an agent to fulfill a formal specification ASAP. The framework leverages a piece-wise reward function that assigns quantitative semantic reward to traces not satisfying the specification, and a high constant reward to the remaining. Then, it trains an agent with an actor-critic-based algorithm, such as soft actor-critic (SAC), or deep deterministic policy gradient (DDPG). Moreover, we prove that ASAP-Phi produces policies that prioritize fulfilling a specification ASAP. Extensive experiments are run, including ablation studies, on state-of-the-art benchmarks. Results show that our framework succeeds in finding sufficiently fast trajectories for up to 97\% test cases and defeats baselines.

Spatial information is essential in various fields. How to explicitly model according to the spatial location of agents is also very important for the multi-agent problem, especially when the number of agents is changing and the scale is enormous. Inspired by the point cloud task in computer vision, we propose a spatial information extraction structure for multi-agent reinforcement learning in this paper. Agents can effectively share the neighborhood and global information through a spatially encoder-decoder structure. Our method follows the centralized training with decentralized execution (CTDE) paradigm. In addition, our structure can be applied to various existing mainstream reinforcement learning algorithms with minor modifications and can deal with the problem with a variable number of agents. The experiments in several multi-agent scenarios show that the existing methods can get convincing results by adding our spatially explicit architecture.