PyTorchKR Papers

최신 논문

총 1,873편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2409.02270

Reinforcement Learning-enabled Satellite Constellation Reconfiguration and Retasking for Mission-Critical Applications

The development of satellite constellation applications is rapidly advancing due to increasing user demands, reduced operational costs, and technological advancements. However, a significant gap in the existing literature concerns reconfiguration and retasking issues within satellite constellations, which is the primary focus of our research. In this work, we critically assess the impact of satellite failures on constellation performance and the associated task requirements. To facilitate this analysis, we introduce a system modeling approach for GPS satellite constellations, enabling an investigation into performance dynamics and task distribution strategies, particularly in scenarios where satellite failures occur during mission-critical operations. Additionally, we introduce reinforcement learning (RL) techniques, specifically Q-learning, Policy Gradient, Deep Q-Network (DQN), and Proximal Policy Optimization (PPO), for managing satellite constellations, addressing the challenges posed by reconfiguration and retasking following satellite failures. Our results demonstrate that DQN and PPO achieve effective outcomes in terms of average rewards, task completion rates, and response times.

arXiv:2409.01369

Imitating Language via Scalable Inverse Reinforcement Learning

The majority of language model training builds on imitation learning. It covers pretraining, supervised fine-tuning, and affects the starting conditions for reinforcement learning from human feedback (RLHF). The simplicity and scalability of maximum likelihood estimation (MLE) for next token prediction led to its role as predominant paradigm. However, the broader field of imitation learning can more effectively utilize the sequential structure underlying autoregressive generation. We focus on investigating the inverse reinforcement learning (IRL) perspective to imitation, extracting rewards and directly optimizing sequences instead of individual token likelihoods and evaluate its benefits for fine-tuning large language models. We provide a new angle, reformulating inverse soft-Q-learning as a temporal difference regularized extension of MLE. This creates a principled connection between MLE and IRL and allows trading off added complexity with increased performance and diversity of generations in the supervised fine-tuning (SFT) setting. We find clear advantages for IRL-based imitation, in particular for retaining diversity while maximizing task performance, rendering IRL a strong alternative on fixed SFT datasets even without online data generation. Our analysis of IRL-extracted reward functions further indicates benefits for more robust reward functions via tighter integration of supervised and preference-based LLM post-training.

arXiv:2408.16882

Coverage Analysis of Multi-Environment Q-Learning Algorithms for Wireless Network Optimization

Q-learning is widely used to optimize wireless networks with unknown system dynamics. Recent advancements include ensemble multi-environment hybrid Q-learning algorithms, which utilize multiple Q-learning algorithms across structurally related but distinct Markovian environments and outperform existing Q-learning algorithms in terms of accuracy and complexity in large-scale wireless networks. We herein conduct a comprehensive coverage analysis to ensure optimal data coverage conditions for these algorithms. Initially, we establish upper bounds on the expectation and variance of different coverage coefficients. Leveraging these bounds, we present an algorithm for efficient initialization of these algorithms. We test our algorithm on two distinct real-world wireless networks. Numerical simulations show that our algorithm can achieve %50 less policy error and %40 less runtime complexity than state-of-the-art reinforcement learning algorithms. Furthermore, our algorithm exhibits robustness to changes in network settings and parameters. We also numerically validate our theoretical results.

arXiv:2408.16981

The Sample-Communication Complexity Trade-off in Federated Q-Learning

We consider the problem of federated Q-learning, where $M$ agents aim to collaboratively learn the optimal Q-function of an unknown infinite-horizon Markov decision process with finite state and action spaces. We investigate the trade-off between sample and communication complexities for the widely used class of intermittent communication algorithms. We first establish the converse result, where it is shown that a federated Q-learning algorithm that offers any speedup with respect to the number of agents in the per-agent sample complexity needs to incur a communication cost of at least an order of $\frac{1}{1-\gamma}$ up to logarithmic factors, where $\gamma$ is the discount factor. We also propose a new algorithm, called Fed-DVR-Q, which is the first federated Q-learning algorithm to simultaneously achieve order-optimal sample and communication complexities. Thus, together these results provide a complete characterization of the sample-communication complexity trade-off in federated Q-learning.

arXiv:2408.16262

On Convergence of Average-Reward Q-Learning in Weakly Communicating Markov Decision Processes

This paper analyzes reinforcement learning (RL) algorithms for Markov decision processes (MDPs) under the average-reward criterion. We focus on Q-learning algorithms based on relative value iteration (RVI), which are model-free stochastic analogues of the classical RVI method for average-reward MDPs. These algorithms have low per-iteration complexity, making them well-suited for large state space problems. We extend the almost-sure convergence analysis of RVI Q-learning algorithms developed by Abounadi, Bertsekas, and Borkar (2001) from unichain to weakly communicating MDPs. This extension is important both practically and theoretically: weakly communicating MDPs cover a much broader range of applications compared to unichain MDPs, and their optimality equations have a richer solution structure (with multiple degrees of freedom), introducing additional complexity in proving algorithmic convergence. We also characterize the sets to which RVI Q-learning algorithms converge, showing that they are compact, connected, potentially nonconvex, and comprised of solutions to the average-reward optimality equation, with exactly one less degree of freedom than the general solution set of this equation. Furthermore, we extend our analysis to two RVI-based hierarchical average-reward RL algorithms using the options framework, proving their almost-sure convergence and characterizing their sets of convergence under the assumption that the underlying semi-Markov decision process is weakly communicating.

arXiv:2408.14864

Dynamic operator management in meta-heuristics using reinforcement learning: an application to permutation flowshop scheduling problems

This study develops a framework based on reinforcement learning to dynamically manage a large portfolio of search operators within meta-heuristics. Using the idea of tabu search, the framework allows for continuous adaptation by temporarily excluding less efficient operators and updating the portfolio composition during the search. A Q-learning-based adaptive operator selection mechanism is used to select the most suitable operator from the dynamically updated portfolio at each stage. Unlike traditional approaches, the proposed framework requires no input from the experts regarding the search operators, allowing domain-specific non-experts to effectively use the framework. The performance of the proposed framework is analyzed through an application to the permutation flowshop scheduling problem. The results demonstrate the superior performance of the proposed framework against state-of-the-art algorithms in terms of optimality gap and convergence speed.

arXiv:2408.14009

Optimizing TD3 for 7-DOF Robotic Arm Grasping: Overcoming Suboptimality with Exploration-Enhanced Contrastive Learning

In actor-critic-based reinforcement learning algorithms such as Twin Delayed Deep Deterministic policy gradient (TD3), insufficient exploration of the spatial space can result in suboptimal policies when controlling 7-DOF robotic arms. To address this issue, we propose a novel Exploration-Enhanced Contrastive Learning (EECL) module that improves exploration by providing additional rewards for encountering novel states. Our module stores previously explored states in a buffer and identifies new states by comparing them with historical data using Euclidean distance within a K-dimensional tree (KDTree) framework. When the agent explores new states, exploration rewards are assigned. These rewards are then integrated into the TD3 algorithm, ensuring that the Q-learning process incorporates these signals, promoting more effective strategy optimization. We evaluate our method on the robosuite panda lift task, demonstrating that it significantly outperforms the baseline TD3 in terms of both efficiency and convergence speed in the tested environment.

arXiv:2408.13184

Can LLM be a Good Path Planner based on Prompt Engineering? Mitigating the Hallucination for Path Planning

Spatial reasoning in Large Language Models (LLMs) is the foundation for embodied intelligence. However, even in simple maze environments, LLMs still encounter challenges in long-term path-planning, primarily influenced by their spatial hallucination and context inconsistency hallucination by long-term reasoning. To address this challenge, this study proposes an innovative model, Spatial-to-Relational Transformation and Curriculum Q-Learning (S2RCQL). To address the spatial hallucination of LLMs, we propose the Spatial-to-Relational approach, which transforms spatial prompts into entity relations and paths representing entity relation chains. This approach fully taps the potential of LLMs in terms of sequential thinking. As a result, we design a path-planning algorithm based on Q-learning to mitigate the context inconsistency hallucination, which enhances the reasoning ability of LLMs. Using the Q-value of state-action as auxiliary information for prompts, we correct the hallucinations of LLMs, thereby guiding LLMs to learn the optimal path. Finally, we propose a reverse curriculum learning technique based on LLMs to further mitigate the context inconsistency hallucination. LLMs can rapidly accumulate successful experiences by reducing task difficulty and leveraging them to tackle more complex tasks. We performed comprehensive experiments based on Baidu's self-developed LLM: ERNIE-Bot 4.0. The results showed that our S2RCQL achieved a 23%--40% improvement in both success and optimality rates compared with advanced prompt engineering.

arXiv:2408.11773

Deviations from the Nash equilibrium and emergence of tacit collusion in a two-player optimal execution game with reinforcement learning

The use of reinforcement learning algorithms in financial trading is becoming increasingly prevalent. However, the autonomous nature of these algorithms can lead to unexpected outcomes that deviate from traditional game-theoretical predictions and may even destabilize markets. In this study, we examine a scenario in which two autonomous agents, modeled with Double Deep Q-Learning, learn to liquidate the same asset optimally in the presence of market impact, using the Almgren-Chriss (2000) framework. Our results show that the strategies learned by the agents deviate significantly from the Nash equilibrium of the corresponding market impact game. Notably, the learned strategies exhibit tacit collusion, closely aligning with the Pareto-optimal solution. We further explore how different levels of market volatility influence the agents' performance and the equilibria they discover, including scenarios where volatility differs between the training and testing phases.

arXiv:2408.09882

GINO-Q: Learning an Asymptotically Optimal Index Policy for Restless Multi-armed Bandits

The restless multi-armed bandit (RMAB) framework is a popular model with applications across a wide variety of fields. However, its solution is hindered by the exponentially growing state space (with respect to the number of arms) and the combinatorial action space, making traditional reinforcement learning methods infeasible for large-scale instances. In this paper, we propose GINO-Q, a three-timescale stochastic approximation algorithm designed to learn an asymptotically optimal index policy for RMABs. GINO-Q mitigates the curse of dimensionality by decomposing the RMAB into a series of subproblems, each with the same dimension as a single arm, ensuring that complexity increases linearly with the number of arms. Unlike recently developed Whittle-index-based algorithms, GINO-Q does not require RMABs to be indexable, enhancing its flexibility and applicability. Our experimental results demonstrate that GINO-Q consistently learns near-optimal policies, even for non-indexable RMABs where Whittle-index-based algorithms perform poorly, and it converges significantly faster than existing baselines.

arXiv:2408.09109

Improved Q-learning based Multi-hop Routing for UAV-Assisted Communication

Designing effective Unmanned Aerial Vehicle(UAV)-assisted routing protocols is challenging due to changing topology, limited battery capacity, and the dynamic nature of communication environments. Current protocols prioritize optimizing individual network parameters, overlooking the necessity for a nuanced approach in scenarios with intermittent connectivity, fluctuating signal strength, and varying network densities, ultimately failing to address aerial network requirements comprehensively. This paper proposes a novel, Improved Q-learning-based Multi-hop Routing (IQMR) algorithm for optimal UAV-assisted communication systems. Using Q(\lambda) learning for routing decisions, IQMR substantially enhances energy efficiency and network data throughput. IQMR improves system resilience by prioritizing reliable connectivity and inter-UAV collision avoidance while integrating real-time network status information, all in the absence of predefined UAV path planning, thus ensuring dynamic adaptability to evolving network conditions. The results validate IQMR's adaptability to changing system conditions and superiority over the current techniques. IQMR showcases 36.35\% and 32.05\% improvements in energy efficiency and data throughput over the existing methods.

arXiv:2408.08242

A Conflicts-free, Speed-lossless KAN-based Reinforcement Learning Decision System for Interactive Driving in Roundabouts

Safety and efficiency are crucial for autonomous driving in roundabouts, especially in the context of mixed traffic where autonomous vehicles (AVs) and human-driven vehicles coexist. This paper introduces a learning-based algorithm tailored to foster safe and efficient driving behaviors across varying levels of traffic flows in roundabouts. The proposed algorithm employs a deep Q-learning network to effectively learn safe and efficient driving strategies in complex multi-vehicle roundabouts. Additionally, a KAN (Kolmogorov-Arnold network) enhances the AVs' ability to learn their surroundings robustly and precisely. An action inspector is integrated to replace dangerous actions to avoid collisions when the AV interacts with the environment, and a route planner is proposed to enhance the driving efficiency and safety of the AVs. Moreover, a model predictive control is adopted to ensure stability and precision of the driving actions. The results show that our proposed system consistently achieves safe and efficient driving whilst maintaining a stable training process, as evidenced by the smooth convergence of the reward function and the low variance in the training curves across various traffic flows. Compared to state-of-the-art benchmarks, the proposed algorithm achieves a lower number of collisions and reduced travel time to destination.

arXiv:2408.06544

Variance-Reduced Cascade Q-learning: Algorithms and Sample Complexity

We study the problem of estimating the optimal Q-function of $\gamma$-discounted Markov decision processes (MDPs) under the synchronous setting, where independent samples for all state-action pairs are drawn from a generative model at each iteration. We introduce and analyze a novel model-free algorithm called Variance-Reduced Cascade Q-learning (VRCQ). VRCQ comprises two key building blocks: (i) the established direct variance reduction technique and (ii) our proposed variance reduction scheme, Cascade Q-learning. By leveraging these techniques, VRCQ provides superior guarantees in the $\ell_\infty$-norm compared with the existing model-free stochastic approximation-type algorithms. Specifically, we demonstrate that VRCQ is minimax optimal. Additionally, when the action set is a singleton (so that the Q-learning problem reduces to policy evaluation), it achieves non-asymptotic instance optimality while requiring the minimum number of samples theoretically possible. Our theoretical results and their practical implications are supported by numerical experiments.

arXiv:2408.04911

A Geometric Nash Approach in Tuning the Learning Rate in Q-Learning Algorithm

This paper proposes a geometric approach for estimating the $\alpha$ value in Q learning. We establish a systematic framework that optimizes the {\alpha} parameter, thereby enhancing learning efficiency and stability. Our results show that there is a relationship between the learning rate and the angle between a vector T (total time steps in each episode of learning) and R (the reward vector for each episode). The concept of angular bisector between vectors T and R and Nash Equilibrium provide insight into estimating $\alpha$ such that the algorithm minimizes losses arising from exploration-exploitation trade-off.

arXiv:2408.04463공식 코드 · Papers with Code

Crowd Intelligence for Early Misinformation Prediction on Social Media

Misinformation spreads rapidly on social media, causing serious damage by influencing public opinion, promoting dangerous behavior, or eroding trust in reliable sources. It spreads too fast for traditional fact-checking, stressing the need for predictive methods. We introduce CROWDSHIELD, a crowd intelligence-based method for early misinformation prediction. We hypothesize that the crowd's reactions to misinformation reveal its accuracy. Furthermore, we hinge upon exaggerated assertions/claims and replies with particular positions/stances on the source post within a conversation thread. We employ Q-learning to capture the two dimensions -- stances and claims. We utilize deep Q-learning due to its proficiency in navigating complex decision spaces and effectively learning network properties. Additionally, we use a transformer-based encoder to develop a comprehensive understanding of both content and context. This multifaceted approach helps ensure the model pays attention to user interaction and stays anchored in the communication's content. We propose MIST, a manually annotated misinformation detection Twitter corpus comprising nearly 200 conversation threads with more than 14K replies. In experiments, CROWDSHIELD outperformed ten baseline systems, achieving an improvement of ~4% macro-F1 score. We conduct an ablation study and error analysis to validate our proposed model's performance. The source code and dataset are available at https://github.com/LCS2-IIITD/CrowdShield.git.

arXiv:2408.03329

Coverage-aware and Reinforcement Learning Using Multi-agent Approach for HD Map QoS in a Realistic Environment

One effective way to optimize the offloading process is by minimizing the transmission time. This is particularly true in a Vehicular Adhoc Network (VANET) where vehicles frequently download and upload High-definition (HD) map data which requires constant updates. This implies that latency and throughput requirements must be guaranteed by the wireless system. To achieve this, adjustable contention windows (CW) allocation strategies in the standard IEEE802.11p have been explored by numerous researchers. Nevertheless, their implementations demand alterations to the existing standard which is not always desirable. To address this issue, we proposed a Q-Learning algorithm that operates at the application layer. Moreover, it could be deployed in any wireless network thereby mitigating the compatibility issues. The solution has demonstrated a better network performance with relatively fewer optimization requirements as compared to the Deep Q Network (DQN) and Actor-Critic algorithms. The same is observed while evaluating the model in a multi-agent setup showing higher performance compared to the single-agent setup.

arXiv:2408.03077

Model-free optimal controller for discrete-time Markovian jump linear systems: A Q-learning approach

This research paper introduces a model-free optimal controller for discrete-time Markovian jump linear systems (MJLSs), employing principles from the methodology of reinforcement learning (RL). While Q-learning methods have demonstrated efficacy in determining optimal controller gains for deterministic systems, their application to systems with Markovian switching remains unexplored. To address this research gap, we propose a Q-function involving the Markovian mode. Subsequently, a Q-learning algorithm is proposed to learn the unknown kernel matrix using raw input-state information from the system. Notably, the study proves the convergence of the proposed Q-learning optimal controller gains to the model-based optimal controller gains after proving the convergence of a value iteration algorithm as the first step. Addition of excitation noise to input which is required to ensure the leaning performance does not lead to any bias. Unlike the conventional optimal controller, the proposed method does not require any knowledge on system dynamics and eliminates the need for solving coupled algebraic Riccati equations arising in optimal control of MJLSs. Finally, the efficiency of the proposed method is demonstrated through a simulation study.

arXiv:2407.19851

Evolution of cooperation in the public goods game with Q-learning

Recent paradigm shifts from imitation learning to reinforcement learning (RL) is shown to be productive in understanding human behaviors. In the RL paradigm, individuals search for optimal strategies through interaction with the environment to make decisions. This implies that gathering, processing, and utilizing information from their surroundings are crucial. However, existing studies typically study pairwise games such as the prisoners' dilemma and employ a self-regarding setup, where individuals play against one opponent based solely on their own strategies, neglecting the environmental information. In this work, we investigate the evolution of cooperation with the multiplayer game -- the public goods game using the Q-learning algorithm by leveraging the environmental information. Specifically, the decision-making of players is based upon the cooperation information in their neighborhood. Our results show that cooperation is more likely to emerge compared to the case of imitation learning by using Fermi rule. Of particular interest is the observation of an anomalous non-monotonic dependence which is revealed when voluntary participation is further introduced. The analysis of the Q-table explains the mechanisms behind the cooperation evolution. Our findings indicate the fundamental role of environment information in the RL paradigm to understand the evolution of cooperation, and human behaviors in general.

arXiv:2407.19634

Evolution of cooperation with Q-learning: the impact of information perception

The inherent complexity of human beings manifests in a remarkable diversity of responses to intricate environments, enabling us to approach problems from varied perspectives. However, in the study of cooperation, existing research within the reinforcement learning framework often assumes that individuals have access to identical information when making decisions, which contrasts with the reality that individuals frequently perceive information differently. In this study, we employ the Q-learning algorithm to explore the impact of information perception on the evolution of cooperation in a two-person Prisoner's Dilemma game. We demonstrate that the evolutionary processes differ significantly across three distinct information perception scenarios, highlighting the critical role of information structure in the emergence of cooperation. Notably, the asymmetric information scenario reveals a complex dynamical process, including the emergence, breakdown, and reconstruction of cooperation, mirroring psychological shifts observed in human behavior. Our findings underscore the importance of information structure in fostering cooperation, offering new insights into the establishment of stable cooperative relationships among humans.

arXiv:2408.03088

QADQN: Quantum Attention Deep Q-Network for Financial Market Prediction

Financial market prediction and optimal trading strategy development remain challenging due to market complexity and volatility. Our research in quantum finance and reinforcement learning for decision-making demonstrates the approach of quantum-classical hybrid algorithms to tackling real-world financial challenges. In this respect, we corroborate the concept with rigorous backtesting and validate the framework's performance under realistic market conditions, by including fixed transaction cost per trade. This paper introduces a Quantum Attention Deep Q-Network (QADQN) approach to address these challenges through quantum-enhanced reinforcement learning. Our QADQN architecture uses a variational quantum circuit inside a traditional deep Q-learning framework to take advantage of possible quantum advantages in decision-making. We gauge the QADQN agent's performance on historical data from major market indices, including the S&P 500. We evaluate the agent's learning process by examining its reward accumulation and the effectiveness of its experience replay mechanism. Our empirical results demonstrate the QADQN's superior performance, achieving better risk-adjusted returns with Sortino ratios of 1.28 and 1.19 for non-overlapping and overlapping test periods respectively, indicating effective downside risk management.