Multi-Agent Deep Reinforcement Learning (MDRL) is a promising research area in which agents learn complex behaviors in cooperative or competitive environments. However, MDRL comes with several challenges that hinder its usability, including sample efficiency, curse of dimensionality, and environment exploration. Recent works proposing Federated Reinforcement Learning (FRL) to tackle these issues suffer from problems related to model restrictions and maliciousness. Other proposals using reward shaping require considerable engineering and could lead to local optima. In this paper, we propose a novel Blockchain-assisted Multi-Expert Demonstration Cloning (MEDC) framework for MDRL. The proposed method utilizes expert demonstrations in guiding the learning of new MDRL agents, by suggesting exploration actions in the environment. A model sharing framework on Blockchain is designed to allow users to share their trained models, which can be allocated as expert models to requesting users to aid in training MDRL systems. A Consortium Blockchain is adopted to enable traceable and autonomous execution without the need for a single trusted entity. Smart Contracts are designed to manage users and models allocation, which are shared using IPFS. The proposed framework is tested on several applications, and is benchmarked against existing methods in FRL, Reward Shaping, and Imitation Learning-assisted RL. The results show the outperformance of the proposed framework in terms of learning speed and resiliency to faulty and malicious models.
The Safe Trusted Autonomy for Responsible Space (STARS) program aims to advance autonomy technologies for space by leveraging machine learning technologies while mitigating barriers to trust, such as uncertainty, opaqueness, brittleness, and inflexibility. This paper presents the achievements and lessons learned from the STARS program in integrating reinforcement learning-based multi-satellite control, run time assurance approaches, and flexible human-autonomy teaming interfaces, into a new integrated testing environment for collaborative autonomous satellite systems. The primary results describe analysis of the reinforcement learning multi-satellite control and run time assurance algorithms. These algorithms are integrated into a prototype human-autonomy interface using best practices from human-autonomy trust literature, however detailed analysis of the effectiveness is left to future work. References are provided with additional detailed results of individual experiments.
Posterior Sampling for Reinforcement Learning (PSRL) is a well-known algorithm that augments model-based reinforcement learning (MBRL) algorithms with Thompson sampling. PSRL maintains posterior distributions of the environment transition dynamics and the reward function, which are intractable for tasks with high-dimensional state and action spaces. Recent works show that dropout, used in conjunction with neural networks, induces variational distributions that can approximate these posteriors. In this paper, we propose Event-based Variational Distributions for Exploration (EVaDE), which are variational distributions that are useful for MBRL, especially when the underlying domain is object-based. We leverage the general domain knowledge of object-based domains to design three types of event-based convolutional layers to direct exploration. These layers rely on Gaussian dropouts and are inserted between the layers of the deep neural network model to help facilitate variational Thompson sampling. We empirically show the effectiveness of EVaDE-equipped Simulated Policy Learning (EVaDE-SimPLe) on the 100K Atari game suite.
Searching for the Extreme Operating Conditions (EOCs) is one of the core problems of power system relay protection setting calculation. The current methods based on brute-force search, heuristic algorithms, and mathematical programming can hardly meet the requirements of today's power systems in terms of computation speed due to the drastic changes in operating conditions induced by renewables and power electronics. This paper proposes an EOC fast search method, named Graph Dueling Double Deep Q Network (Graph D3QN), which combines graph neural network and deep reinforcement learning to address this challenge. First, the EOC search problem is modeled as a Markov decision process, where the information of the underlying power system is extracted using graph neural networks, so that the EOC of the system can be found via deep reinforcement learning. Then, a two-stage Guided Learning and Free Exploration (GLFE) training framework is constructed to accelerate the convergence speed of reinforcement learning. Finally, the proposed Graph D3QN method is validated through case studies of searching maximum fault current for relay protection setting calculation on the IEEE 39-bus and 118-bus systems. The experimental results demonstrate that Graph D3QN can reduce the computation time by 10 to 1000 times while guaranteeing the accuracy of the selected EOCs.
In reinforcement learning, the value function is typically trained to solve the Bellman equation, which connects the current value to future values. This temporal dependency hints that the value function may contain implicit information about the environment's transition dynamics. By rearranging the Bellman equation, we show that a converged value function encodes a model of the underlying dynamics of the environment. We build on this insight to propose a simple method for inferring dynamics models directly from the value function, potentially mitigating the need for explicit model learning. Furthermore, we explore the challenges of next-state identifiability, discussing conditions under which the inferred dynamics model is well-defined. Our work provides a theoretical foundation for leveraging value functions in dynamics modeling and opens a new avenue for bridging model-free and model-based reinforcement learning.
The average-reward formulation of reinforcement learning (RL) has drawn increased interest in recent years due to its ability to solve temporally-extended problems without discounting. Independently, RL algorithms have benefited from entropy-regularization: an approach used to make the optimal policy stochastic, thereby more robust to noise. Despite the distinct benefits of the two approaches, the combination of entropy regularization with an average-reward objective is not well-studied in the literature and there has been limited development of algorithms for this setting. To address this gap in the field, we develop algorithms for solving entropy-regularized average-reward RL problems with function approximation. We experimentally validate our method, comparing it with existing algorithms on standard benchmarks for RL.
Urban air mobility (UAM) is a transformative system that operates various small aerial vehicles in urban environments to reshape urban transportation. However, integrating UAM into existing urban environments presents a variety of complex challenges. Recent analyses of UAM's operational constraints highlight aircraft noise and system safety as key hurdles to UAM system implementation. Future UAM air traffic management schemes must ensure that the system is both quiet and safe. We propose a multi-agent reinforcement learning approach to manage UAM traffic, aiming at both vertical separation assurance and noise mitigation. Through extensive training, the reinforcement learning agent learns to balance the two primary objectives by employing altitude adjustments in a multi-layer UAM network. The results reveal the tradeoffs among noise impact, traffic congestion, and separation. Overall, our findings demonstrate the potential of reinforcement learning in mitigating UAM's noise impact while maintaining safe separation using altitude adjustments
This paper summarizes in depth the state of the art of aerial swarms, covering both classical and new reinforcement-learning-based approaches for their management. Then, it proposes a hybrid AI system, integrating deep reinforcement learning in a multi-agent centralized swarm architecture. The proposed system is tailored to perform surveillance of a specific area, searching and tracking ground targets, for security and law enforcement applications. The swarm is governed by a central swarm controller responsible for distributing different search and tracking tasks among the cooperating UAVs. Each UAV agent is then controlled by a collection of cooperative sub-agents, whose behaviors have been trained using different deep reinforcement learning models, tailored for the different task types proposed by the swarm controller. More specifically, proximal policy optimization (PPO) algorithms were used to train the agents' behavior. In addition, several metrics to assess the performance of the swarm in this application were defined. The results obtained through simulation show that our system searches the operation area effectively, acquires the targets in a reasonable time, and is capable of tracking them continuously and consistently.
In this paper, we propose an Adaptive Neuro-Symbolic Learning and Reasoning Framework for digital twin technology called ``ANSR-DT." Digital twins in industrial environments often struggle with interpretability, real-time adaptation, and human input integration. Our approach addresses these challenges by combining CNN-LSTM dynamic event detection with reinforcement learning and symbolic reasoning to enable adaptive intelligence with interpretable decision processes. This integration enhances environmental understanding while promoting continuous learning, leading to more effective real-time decision-making in human-machine collaborative applications. We evaluated ANSR-DT on synthetic industrial data, observing significant improvements over traditional approaches, with up to 99.5% accuracy for dynamic pattern recognition. The framework demonstrated superior adaptability with extended reinforcement learning training, improving explained variance from 0.447 to 0.547. Future work aims at scaling to larger datasets to test rule management beyond the current 14 rules. Our open-source implementation promotes reproducibility and establishes a foundation for future research in adaptive, interpretable digital twins for industrial applications.
Procedural Content Generation (PCG) is widely used to create scalable and diverse environments in games. However, existing methods, such as the Wave Function Collapse (WFC) algorithm, are often limited to static scenarios and lack the adaptability required for dynamic, narrative-driven applications, particularly in augmented reality (AR) games. This paper presents a reinforcement learning-enhanced WFC framework designed for mobile AR environments. By integrating environment-specific rules and dynamic tile weight adjustments informed by reinforcement learning (RL), the proposed method generates maps that are both contextually coherent and responsive to gameplay needs. Comparative evaluations and user studies demonstrate that the framework achieves superior map quality and delivers immersive experiences, making it well-suited for narrative-driven AR games. Additionally, the method holds promise for broader applications in education, simulation training, and immersive extended reality (XR) experiences, where dynamic and adaptive environments are critical.
Large language models (LLMs) are remarked by their substantial computational requirements. To mitigate the cost, researchers develop specialized CUDA kernels, which often fuse several tensor operations to maximize the utilization of GPUs as much as possible. However, those specialized kernels may still leave performance on the table as CUDA assembly experts show that manual optimization of GPU SASS schedules can lead to better performance, and trial-and-error is largely employed to manually find the best GPU SASS schedules. In this work, we employ an automatic approach to optimize GPU SASS schedules, which thus can be integrated into existing compiler frameworks. The key to automatic optimization is training an RL agent to mimic how human experts perform manual scheduling. To this end, we formulate an assembly game, where RL agents can play to find the best GPU SASS schedules. The assembly game starts from a \textit{-O3} optimized SASS schedule, and the RL agents can iteratively apply actions to mutate the current schedules. Positive rewards are generated if the mutated schedules get higher throughput by executing on GPUs. Experiments show that CuAsmRL can further improve the performance of existing specialized CUDA kernels transparently by up to $26\%$, and on average $9\%$. Moreover, it is used as a tool to reveal potential optimization moves learned automatically.
This paper explores multiple optimization methods to improve the performance of rating-based reinforcement learning (RbRL). RbRL, a method based on the idea of human ratings, has been developed to infer reward functions in reward-free environments for the subsequent policy learning via standard reinforcement learning, which requires the availability of reward functions. Specifically, RbRL minimizes the cross entropy loss that quantifies the differences between human ratings and estimated ratings derived from the inferred reward. Hence, a low loss means a high degree of consistency between human ratings and estimated ratings. Despite its simple form, RbRL has various hyperparameters and can be sensitive to various factors. Therefore, it is critical to provide comprehensive experiments to understand the impact of various hyperparameters on the performance of RbRL. This paper is a work in progress, providing users some general guidelines on how to select hyperparameters in RbRL.
This paper addresses a critical challenge in the high-speed passenger railway industry: designing effective dynamic pricing strategies in the context of competing and cooperating operators. To address this, a multi-agent reinforcement learning (MARL) framework based on a non-zero-sum Markov game is proposed, incorporating random utility models to capture passenger decision making. Unlike prior studies in areas such as energy, airlines, and mobile networks, dynamic pricing for railway systems using deep reinforcement learning has received limited attention. A key contribution of this paper is a parametrisable and versatile reinforcement learning simulator designed to model a variety of railway network configurations and demand patterns while enabling realistic, microscopic modelling of user behaviour, called RailPricing-RL. This environment supports the proposed MARL framework, which models heterogeneous agents competing to maximise individual profits while fostering cooperative behaviour to synchronise connecting services. Experimental results validate the framework, demonstrating how user preferences affect MARL performance and how pricing policies influence passenger choices, utility, and overall system dynamics. This study provides a foundation for advancing dynamic pricing strategies in railway systems, aligning profitability with system-wide efficiency, and supporting future research on optimising pricing policies.
Satellite communication is a key technology in our modern connected world. With increasingly complex hardware, one challenge is to efficiently configure links (connections) on a satellite transponder. Planning an optimal link configuration is extremely complex and depends on many parameters and metrics. The optimal use of the limited resources, bandwidth and power of the transponder is crucial. Such an optimization problem can be approximated using metaheuristic methods such as simulated annealing, but recent research results also show that reinforcement learning can achieve comparable or even better performance in optimization methods. However, there have not yet been any studies on link configuration on satellite transponders. In order to close this research gap, a transponder environment was developed as part of this work. For this environment, the performance of the reinforcement learning algorithm PPO was compared with the metaheuristic simulated annealing in two experiments. The results show that Simulated Annealing delivers better results for this static problem than the PPO algorithm, however, the research in turn also underlines the potential of reinforcement learning for optimization problems.
Pre-trained transformer models such as BERT have shown massive gains across many text classification tasks. However, these models usually need enormous labeled data to achieve impressive performances. Obtaining labeled data is often expensive and time-consuming, whereas collecting unlabeled data using some heuristics is relatively much cheaper for any task. Therefore, this paper proposes a method that encapsulates reinforcement learning-based text generation and semi-supervised adversarial learning approaches in a novel way to improve the model's performance. Our method READ, Reinforcement-based Adversarial learning, utilizes an unlabeled dataset to generate diverse synthetic text through reinforcement learning, improving the model's generalization capability using adversarial learning. Our experimental results show that READ outperforms the existing state-of-art methods on multiple datasets.
Multi-objective decision-making problems have emerged in numerous real-world scenarios, such as video games, navigation and robotics. Considering the clear advantages of Reinforcement Learning (RL) in optimizing decision-making processes, researchers have delved into the development of Multi-Objective RL (MORL) methods for solving multi-objective decision problems. However, previous methods either cannot obtain the entire Pareto front, or employ only a single policy network for all the preferences over multiple objectives, which may not produce personalized solutions for each preference. To address these limitations, we propose a novel decomposition-based framework for MORL, Pareto Set Learning for MORL (PSL-MORL), that harnesses the generation capability of hypernetwork to produce the parameters of the policy network for each decomposition weight, generating relatively distinct policies for various scalarized subproblems with high efficiency. PSL-MORL is a general framework, which is compatible for any RL algorithm. The theoretical result guarantees the superiority of the model capacity of PSL-MORL and the optimality of the obtained policy network. Through extensive experiments on diverse benchmarks, we demonstrate the effectiveness of PSL-MORL in achieving dense coverage of the Pareto front, significantly outperforming state-of-the-art MORL methods in the hypervolume and sparsity indicators.
In this paper, we address a crucial but often overlooked issue in applying reinforcement learning (RL) to radio resource management (RRM) in wireless communications: the mismatch between the discounted reward RL formulation and the undiscounted goal of wireless network optimization. To the best of our knowledge, we are the first to systematically investigate this discrepancy, starting with a discussion of the problem formulation followed by simulations that quantify the extent of the gap. To bridge this gap, we introduce the use of average reward RL, a method that aligns more closely with the long-term objectives of RRM. We propose a new method called the Average Reward Off policy Soft Actor Critic (ARO SAC) is an adaptation of the well known Soft Actor Critic algorithm in the average reward framework. This new method achieves significant performance improvement our simulation results demonstrate a 15% gain in the system performance over the traditional discounted reward RL approach, underscoring the potential of average reward RL in enhancing the efficiency and effectiveness of wireless network optimization.
When applied in healthcare, reinforcement learning (RL) seeks to dynamically match the right interventions to subjects to maximize population benefit. However, the learned policy may disproportionately allocate efficacious actions to one subpopulation, creating or exacerbating disparities in other socioeconomically-disadvantaged subgroups. These biases tend to occur in multi-stage decision making and can be self-perpetuating, which if unaccounted for could cause serious unintended consequences that limit access to care or treatment benefit. Counterfactual fairness (CF) offers a promising statistical tool grounded in causal inference to formulate and study fairness. In this paper, we propose a general framework for fair sequential decision making. We theoretically characterize the optimal CF policy and prove its stationarity, which greatly simplifies the search for optimal CF policies by leveraging existing RL algorithms. The theory also motivates a sequential data preprocessing algorithm to achieve CF decision making under an additive noise assumption. We prove and then validate our policy learning approach in controlling unfairness and attaining optimal value through simulations. Analysis of a digital health dataset designed to reduce opioid misuse shows that our proposal greatly enhances fair access to counseling.
This paper presents a hierarchical reinforcement learning (RL) approach to address the agent grouping or pairing problem in cooperative multi-agent systems. The goal is to simultaneously learn the optimal grouping and agent policy. By employing a hierarchical RL framework, we distinguish between high-level decisions of grouping and low-level agents' actions. Our approach utilizes the CTDE (Centralized Training with Decentralized Execution) paradigm, ensuring efficient learning and scalable execution. We incorporate permutation-invariant neural networks to handle the homogeneity and cooperation among agents, enabling effective coordination. The option-critic algorithm is adapted to manage the hierarchical decision-making process, allowing for dynamic and optimal policy adjustments.
This paper presents a novel approach combining inductive logic programming with reinforcement learning to improve training performance and explainability. We exploit inductive learning of answer set programs from noisy examples to learn a set of logical rules representing an explainable approximation of the agent policy at each batch of experience. We then perform answer set reasoning on the learned rules to guide the exploration of the learning agent at the next batch, without requiring inefficient reward shaping and preserving optimality with soft bias. The entire procedure is conducted during the online execution of the reinforcement learning algorithm. We preliminarily validate the efficacy of our approach by integrating it into the Q-learning algorithm for the Pac-Man scenario in two maps of increasing complexity. Our methodology produces a significant boost in the discounted return achieved by the agent, even in the first batches of training. Moreover, inductive learning does not compromise the computational time required by Q-learning and learned rules quickly converge to an explanation of the agent policy.